Opinion

In this action, we determine the sufficiency of an environmental impact report (EIR) on the proposed relocation within the City of San Francisco of biomedical research facilities of the School of Pharmacy at the University of California, San Francisco (UCSF).

A neighborhood association challenges the EIR on three primary grounds; (1) that it does not discuss anticipated future activities at the new location and the effects of those activities; (2) that it does not adequately discuss feasible alternatives to the project; and (3) that there is no substantial evidence the project’s adverse environmental effects will be mitigated. These challenges are based on the California Environmental Quality Act (CEQA)(Pub. Resources Code, § 21000 et seq.).

We find the EIR was inadequate because: (1) it fails to discuss the anticipated future uses of the new facility and the environmental effects of those uses, and (2) the discussion of alternatives is inadequate under CEQA. We find, however, there is substantial evidence the environmental effects identified in the present EIR will be sufficiently mitigated. We also find the Court of Appeal properly determined the neighborhood association is entitled to an award of its attorneys fees under Code of Civil Procedure section 1021.5.

Because the EIR is invalid in part, a new EIR must be prepared, submitted for public review and comment, and certified in accord with CEQA procedures. We decline, however, to order UCSF’s present activities at the new location stayed pending certification of a new EIR.

Facts

Many of the relevant facts in this case are of a detailed, technical nature and are best understood in the context of the legal issues presented. We will set forth those facts in our discussion. The following is the background of the dispute we are asked to resolve:

The UCSF Parnassus campus in San Francisco is the site of the University’s Schools of Medicine, Nursing, Pharmacy, and Dentistry. In 1982, the University of California (University) prepared a long range development plan for UCSF, which indicated there were serious space constraints at the Parnassus campus and concluded there was a need to develop off-campus locations for academic and support activities.

To alleviate these space constraints, in February 1985 the Regents of the University of California purchased the Presidio Corporate Center, formerly known as the Fireman’s Fund Insurance Building, located in the Laurel Heights neighborhood of San Francisco, approximately two miles northeast of the Parnassus campus. The Laurel Heights neighborhood is a mixture of residential and commercial development. The facility purchased by the Regents is a 10-acre site containing a 354,000 square-foot building (exclusive of parking area) and a 13,000 square-foot annex.

The Regents initially claimed the purchase had no significant environmental effects and was exempt from CEQA’s requirement for an EIR because the relocation to Laurel Heights would involve only the acquisition and operation of an existing facility and negligible or no expansion of existing use at that facility. The Regents subsequently decided an EIR was necessary. The reason for the change of opinion appears to have been a decision after the purchase to relocate School of Pharmacy biomedical research units to the Laurel Heights facility.

UCSF prepared a draft EIR, which stated that “The UCSF proposal is to move the School of Pharmacy basic science research units from the UCSF Parnassus campus to Laurel Heights.” The draft EIR also indicated that a secondary objective was to consolidate scattered School of Pharmacy facilities into a single building. The draft EIR disclosed that the basic science research units to be relocated included a number of facilities that handled possibly toxic chemicals, possible carcinogens, and radioactive substances; that various substances would be vented from the building into the outside air; that hazardous wastes would be generated; and that harmful exposure to hazardous substances could occur through worker negligence, accidents, or unidentified risks. Potential environmental effects identified in the draft EIR included direct and cumulative effects on air quality caused by laboratory emissions vented into the outside air and effects on human health from exposure to hazardous chemicals. The draft EIR also identified other effects, including noise, traffic congestion, and parking.

The proposed relocation has provoked an intense and continuing controversy in the Laurel Heights neighborhood. The primary dispute is whether scientific research using toxic chemicals, carcinogens, and radioactive materials is too high-risk to be conducted in a residential neighborhood. After a 45-day period for public review of the EIR and comment, the Regents held a public meeting to respond to comments received during the review period. UCSF proposed measures to mitigate the identified environmental effects and prepared a final EIR, concluding that the environmental effects had been “reduced to a level of insignificance.” The Regents certified the final EIR.

The Laurel Heights Neighborhood Improvement Association, Inc. (Association) then petitioned for a writ of mandate setting aside the EIR approval. The superior court denied the petition and in a written statement of decision concluded the Regents had certified the EIR in the manner required by law and that their action was supported by substantial evidence. The Association appealed.

The Court of Appeal reversed on three primary grounds. First, it found the EIR did not adequately describe the “project” within the meaning of CEQA because the EIR did not discuss the future cumulative effects of the relocation of additional UCSF operations to the Laurel Heights site. Second, the Court of Appeal found inadequate the EIR’s discussion of project alternatives. Third, the court found no substantial evidence to support the Regents’ conclusion that all significant environmental effects will be mitigated.

The Court of Appeal denied the Regents’ petition for a rehearing and, at the same time, granted the Association’s request for attorneys’ fees under Code of Civil Procedure section 1021.5 and issued a stay enjoining the Regents from conducting any research at the Laurel Heights facility for 90 days. We then stayed the Court of Appeal’s order in its entirety, but subsequently modified our order to prohibit the Regents from introducing radioactive materials to the facility pending further order of this court. We then granted the Regents’ petition for review of the Court of Appeal’s decision on the merits.

Discussion

With the exception of the Court of Appeal’s award of attorneys fees to the Association, all the substantive issues before us relate to the sufficiency of the final EIR certified by the Regents and the finding that potential environmental effects will be mitigated to a level of insignificance. An understanding of the purposes of EIR’s and their role in the protection of California’s environmental resources is therefore a necessary foundation for our decision.

The foremost principle under CEQA is that the Legislature intended the act “to be interpreted in such manner as to afford the fullest possible protection to the environment within the reasonable scope of the statutory language.” (Friends of Mammoth v. Board of Supervisors (1972) 8 Cal.3d 247, 259 [ 104 Cal.Rptr. 761 , 502 P.2d 1049 ].) More than a decade ago, we observed that, “It is, of course, too late to argue for a grudging, miserly reading of CEQA.” (Bozung v. Local Agency Formation Com. (1975) 13 Cal.3d 263, 274 [ 118 Cal.Rptr. 249 , 529 P.2d 1017 ] [hereafter Bozung].) The Legislature has emphasized that “It is the intent of the Legislature that all agencies of the state government which regulate activities . . . which are found to affect the quality of the environment, shall regulate such activities so that major consideration is given to preventing environmental damage . . . .” (§ 21000, subd. (g).)

With narrow exceptions, CEQA requires an EIR whenever a public agency proposes to approve or to carry out a project that may have a significant effect on the environment. (§ 21100 [state agencies]; § 21151 [local agencies]; Guidelines, § 15002, subd. (f)(1).) “Project” means, among other things, “[activities directly undertaken by any public agency.” (§21065, subd. (a).) “‘Significant effect on the environment’ means a substantial, or potentially substantial, adverse change in the environment.” (§ 21068; see also Guidelines, § 15002, subd. (g).) The Legislature has made clear that an EIR is “an informational document” and that “[t]he purpose of an environmental impact report is to provide public agencies and the public in general with detailed information about the effect which a proposed project is likely to have on the environment; to list ways in which the significant effects of such a project might be minimized; and to indicate alternatives to such a project.” (§ 21061; Guidelines, § 15003, subds. (b)-

(e)-)

Under CEQA, the public is notified that a draft EIR is being prepared (§§ 21092 and 21092.1), and the draft EIR is evaluated in light of comments received. (Guidelines, §§ 15087 and 15088.) The lead agency then prepares a final EIR incorporating comments on the draft EIR and the agency’s responses to significant environmental points raised in the review process. (Guidelines, §§ 15090 and 15132, subds. (b)-(d).) The lead agency must certify that the final EIR has been completed in compliance with CEQA and that the information in the final EIR was considered by the agency before approving the project. (Guidelines, § 15090.) Before approving the project, the agency must also find either that the project’s significant environmental effects identified in the EIR have been avoided or mitigated, or that unmitigated effects are outweighed by the project’s benefits. (§§ 21002, 21002.1, and 21081; Guidelines, §§ 15091-15093.)

The EIR is the primary means of achieving the Legislature’s considered declaration that it is the policy of this state to “take all action necessary to protect, rehabilitate, and enhance the environmental quality of the state.” (§ 21001, subd. (a).) The EIR is therefore “the heart of CEQA.” (Guidelines, § 15003, subd. (a); County of Inyo v. Yorty (1973) 32 Cal.App.3d 795, 810 [ 108 Cal.Rptr. 377 ].) An EIR is an “environmental ‘alarm bell’ whose purpose it is to alert the public and its responsible officials to environmental changes before they have reached ecological points of no return.” (Ibid.; Santiago County Water Dist. v. County of Orange (1981) 118 Cal.App.3d 818, 822 [173 Cal.ptr. 602].) The EIR is also intended “to demonstrate to an apprehensive citizenry that the agency has, in fact, analyzed and considered the ecological implications of its action.” (No Oil, Inc. v. City of Los Angeles (1974) 13 Cal.3d 68, 86 [ 118 Cal.Rptr. 34 , 529 P.2d 66 ] [hereafter No Oil]; Guidelines, § 15003, subd. (d).) Because the EIR must be certified or rejected by public officials, it is a document of accountability. If CEQA is scrupulously followed, the public will know the basis on which its responsible officials either approve or reject environmentally significant action, and the public, being duly informed, can respond accordingly to action with which it disagrees. (People v. County of Kern (1974) 39 Cal.App.3d 830, 842 [ 115 Cal.Rptr. 67 ]; Guidelines, § 15003, subd. (e).) The EIR process protects not only the environment but also informed self-government.

Section 21168.5 provides that a court’s inquiry in an action to set aside an agency’s decision under CEQA “shall extend only to whether there was a prejudicial abuse of discretion. Abuse of discretion is established if the agency has not proceeded in a manner required by law or if the determination or decision is not supported by substantial evidence.” As a result of this standard, “The court does not pass upon the correctness of the EIR’s environmental conclusions, but only upon its sufficiency as an informative document.” (County of Inyo v. City of Los Angeles (1977) 71 Cal.App.3d 185, 189 [ 139 Cal.Rptr. 396 ].)

This standard of review is consistent with the requirement that the agency’s approval of an EIR “shall be supported by substantial evidence in the record.” (Guidelines, § 15091, subd. (b).) In applying the substantial evidence standard, “the reviewing court must resolve reasonable doubts in favor of the administrative finding and decision.” (Topanga Association for a Scenic Community v. County of Los Angeles (1974) 11 Cal.3d 506, 514 [ 113 Cal.Rptr. 836 , 522 P.2d 12 ].) The Guidelines define “substantial evidence” as “enough relevant information and reasonable inferences from this information that a fair argument can be made to support a conclusion, even though other conclusions might also be reached.” (Guidelines, § 15384, subd. (a).)

A court may not set aside an agency’s approval of an EIR on the ground that an opposite conclusion would have been equally or more reasonable. (Greenebaum v. City of Los Angeles (1984) 153 Cal.App.3d 391, 401-402 [ 200 Cal.Rptr. 237 ].) A court’s task is not to weigh conflicting evidence and determine who has the better argument when the dispute is whether adverse effects have been mitigated or could be better mitigated. We have neither the resources nor scientific expertise to engage in such analysis, even if the statutorily prescribed standard of review permitted us to do so. Our limited function is consistent with the principle that “The purpose of CEQA is not to generate paper, but to compel government at all levels to make decisions with environmental consequences in mind. CEQA does not, indeed cannot, guarantee that these decisions will always be those which favor environmental considerations.” ( Bozung, supra, 13 Cal.3d 263, 283 .).

With the foregoing precepts to guide us, we turn to the issues at hand.

I.

The EIR’s analysis of future activity and effects is inadequate under CEQA.

The EIR before us defined the project as “mov[ing] the School of Pharmacy basic science research units from the UCSF Parnassus campus to Laurel Heights.” The building to which those research units are to be moved is approximately 354,000 square feet in size, but only 100,000 square feet are now available to UCSF because one-half of the building is occupied by the California Department of Transportation (CALTRANS) pursuant to a lease with the University that expires in 1990 with an option to extend tenancy until 1995. (A small portion of the building is leased to private tenants.) The EIR does not discuss the additional environmental effects, if any, that will result from UCSF’s use of the remaining 254,000 square feet when it becomes available, perhaps as soon as 1990.

The Association contends the EIR is inadequate because it fails to discuss the anticipated future uses of the Laurel Heights facility and the likely effects of those uses. The Regents contend they need not evaluate the effects of future uses because the Regents have not yet formally approved any particular use of the remaining space.

“CEQA requires that an agency determine whether a project may have a significant environmental impact, and thus whether an EIR is required, before it approves that project.” (No Oil, supra, 13 Cal.3d 68, 79 , italics by court; Village Laguna of Laguna Beach, Inc. v. Board of Supervisors (1982) 134 Cal.App.3d 1022, 1026 [ 185 Cal.Rptr. 41 ].) This requirement is obvious in several sections of CEQA. For example, section 21081 refers to approval of a project for which an EIR “has been completed,” and section 21151 requires an EIR for a project an agency “intend\s\ to carry out or approve.” (Italics added.) The Guidelines provide even more explicitly that Before granting any approval of a project subject to CEQA, every lead agency . . . shall consider a final EIR . . . .” (Guidelines, § 15004 subd. (a), italics added.) A fundamental purpose of an EIR is to provide decision makers with information they can use in deciding whether to approve a proposed project, not to inform them of the environmental effects of projects that they have already approved. If postapproval environmental review were allowed, EIR’s would likely become nothing more than post hoc rationalizations to support action already taken. We have expressly condemned this use of EIR’s. (No Oil, supra, 13 Cal.3d at p. 79.) The Regents’ view that their approval of a project is the predicate for an EIR stands this principle on its head.

The Regents’ view is also inconsistent with the related rule that significant cumulative effects of a project must be considered in an EIR. (§ 21083, subd. (b); Guidelines, § 15130, subd. (a); Bozung, supra, 13 Cal.3d at pp. 283-284; Environmental Protection Information Center, Inc. v. Johnson (1985) 170 Cal.App.3d 604, 624-625 [ 216 Cal.Rptr. 502 ].) The Guidelines explain that a discussion of cumulative effects should encompass “past, present, and reasonably anticipated future projects.” (Guidelines, § 15130, subd. (b)(1)(A), italics added.)

We hold that a public agency’s approval of a project or future portions of a project is not a prerequisite for an environmental impact report under CEQA.

The more important and difficult question is what circumstances require consideration in an EIR of future action related to the proposed project. A basic tenet of CEQA is that an environmental analysis “should be prepared as early as feasible in the planning process to enable environmental considerations to influence project program and design and yet late enough to provide meaningful information for environmental assessment.” (Guidelines, § 15004, subd. (b); No Oil, supra, 13 Cal.3d 68, 77, fn. 5 .) The Regents correctly note that “where future development is unspecified and uncertain, no purpose can be served by requiring an EIR to engage in sheer speculation as to future environmental consequences.” (Lake County Energy Council v. County of Lake (1977) 70 Cal.App.3d 851, 854-855 [ 139 Cal.Rptr. 176 ].) We agree that environmental resources and the public fisc may be ill served if the environmental review is too early. On the other hand, the later the environmental review process begins, the more bureaucratic and financial momentum there is behind a proposed project, thus providing a strong incentive to ignore environmental concerns that could be dealt with more easily at an early stage of the project. This problem may be exacerbated where, as here, the public agency prepares and approves the EIR for its own project. For that reason, “ ‘EIRs should be prepared as early in the planning process as possible to enable environmental considerations to influence project, program or design.’ ” (Bozung, supra, 13 Cal.3d at p. 282; Guidelines, § 15004, subd. (b).) The University’s own “Procedures for Implementation of the California Environmental Quality Act” state, “. . . in planning for each University project, environmental concerns are taken into account as early as possible ... to influence project program and design.”

The correct answer to the question of how to balance these competing concerns is suggested by our opinion in No Oil, supra, 13 Cal.3d 68 , in which the plaintiffs contended the trial court had erred in limiting the scope of the project at issue to the drilling of two exploratory oil wells and that the project should have been defined to include commercial oil production that would likely commence if the test wells were successful. The defendants argued that geologic information obtained from the two test wells was essential to the preparation of a meaningful EIR on the effect of future commercial production. (Id., at p. 77, fn. 5.) Because we decided the case on other grounds, we did not determine whether the project had been properly defined, but we framed the issue as whether the public agency had “sufficient reliable data to permit preparation of a meaningful and accurate report on the impact of commercial production.” (Ibid.) We did not frame the issue in terms of whether the public agency or the project proponent had any definite plans for action after test drilling.

We hold that an EIR must include an analysis of the environmental effects of future expansion or other action if: (1) it is a reasonably foreseeable consequence of the initial project; and (2) the future expansion or action will be significant in that it will likely change the scope or nature of the initial project or its environmental effects. Absent these two circumstances, the future expansion need not be considered in the EIR for the proposed project. Of course, if the future action is not considered at that time, it will have to be discussed in a subsequent EIR before the future action can be approved under CEQA.

This standard is consistent with the principle that “environmental considerations do not become submerged by chopping a large project into many little ones—each with a minimal potential impact on the environment— which cumulatively may have disastrous consequences.” ( Bozung, supra, 13 Cal.3d at pp. 283-284.) The standard also gives due deference to the fact that premature environmental analysis may be meaningless and financially wasteful. Under this standard, the facts of each case will determine whether and to what extent an EIR must analyze future expansion or other action.

The draft EIR acknowledged that UCSF will occupy the entire Laurel Heights facility when the remainder of the space becomes available. In response to public inquiry as to plans for the facility, UCSF explained that it intends to use the facility for the School of Pharmacy’s basic science group and UCSF’s Office of the Dean. The EIR even estimated the number of faculty, staff, and students that will occupy the facility until 1995 (a total of 460 persons) and then afterward when the entire facility becomes available (860 persons). Under the standard we have announced, it is therefore indisputable that the future expansion and general type of future use is reasonably foreseeable. This is not the type of situation where it is unclear as to whether a parcel of land will be developed or as to whether activity will commence. For example, in No Oil, supra, 13 Cal. 3d 68 , whether commercial oil production would ever occur was entirely speculative. There is no doubt, however, that in this case there will be future use.

The Regents’ contention is only that they have not formally decided precisely how they will use the remainder of the building. That argument is beside the point. They have admitted that they intend to use the entire facility, and, in light of the record before us, it is reasonably foreseeable that the facility will be used primarily for the School of Pharmacy, more specifically, as a biomedical research facility. For example, the draft EIR states that if CALTRANS does not renew its lease in 1990, “the site would be developed as a biomedical research facility, with cross disciplinary programs from all UCSF schools.”

The final EIR contains the following quote from a March 1986 public newsletter by UCSF’s Chancellor: “[A]fter consultation with the other schools, it became clear that with this move [i.e., the present project] the best use of the Laurel Heights site we could make, when it becomes fully available to us in 1995, would be to develop it as a biomedical research facility, with cross-disciplinary programs from all the schools.” (Italics added.) The same newsletter stated that UCSF had made a “final decision” to move the School of Pharmacy to the Laurel Heights site and that UCSF was then “in the midst of completing an Environmental Impact Report (EIR), which will deal not only with the School of Pharmacy component but the long range use of the building.” (Italics added.) The minutes of a May 1986 meeting of UCSF’s Laurel Heights Campus Planning Committee state: “There was a concern that the DEIR [draft EIR] did not discuss the program plans and impacts of the building after the Cal-Trans lease expires. Dean Goyan [School of Pharmacy] confirmed that the building will be dedicated primarily to biomedical research. There are no plans for extensive student activities or clinical activities to be located at the site after 1995.” (Italics added.)

There is more. In addition to these public disclosures, private correspondence makes clear the University’s plan. In a May 1985 letter, the Pharmacy Dean asked the Chancellor for confirmation that the Laurel Heights facility would be committed to basic research. In November 1985, the Chancellor confirmed in writing that “at least 80 percent of the building after total occupancy by UCSF will be devoted to academic units primarily related to biomedical research.” (Italics in original.)

In short, there is telling evidence that the University, by the time it prepared the EIR, had either made decisions or formulated reasonably definite proposals as to future uses of the building. At a minimum, it is clear that the future expansion and the general types of future activity at the facility are reasonably foreseeable.

To counter this evidence the Regents argue that only they can approve formal plans as to the building’s future use and that statements by the Chancellor, Dean, and other officials are insignificant. We need not delve into the University’s complex internal procedures to determine who has the power to decide precise uses of the building. The point is that there is credible and substantial evidence that UCSF’s plans are reasonably foreseeable. It is the substance of the evidence, not the source alone, that matters.

We also find the future action will be significant in that it will likely change the scope or nature of the proposed initial project and its environmental effects. The Regents do not contend otherwise, and could not reasonably do so. The anticipated eventual use of the entire Laurel Heights facility would include an increase in the amount of space used from approximately 100,000 square feet to 354,000 square feet and an increase in occupants from approximately 460 to 860. This is obviously a change in the scope of the project and perhaps its nature as well.

We believe the Regents can provide meaningful, reliable data in the EIR as to future activity at Laurel Heights and thus must do so. A factually similar situation was present in Whitman v. Board of Supervisors (1979) 88 Cal.App.3d 397 [ 151 Cal.Rptr. 866 ]. An oil company’s application for a conditional use permit was granted despite the EIR’s failure to discuss the environmental effects of a contemplated pipeline. The court found the EIR to be inadequate and explained that “The record before us reflects that the construction of the pipeline was, from the very beginning, within the contemplation of [the project proponent] should its well prove productive. Although admittedly contingent on the happening of certain occurrences, the pipeline was, nevertheless part of [the] overall plan for the project and could have been discussed in the EIR in at least general terms.'” (Id., at pp. 414-415 [italics added]; No Oil, Inc. v. City of Los Angeles (1987) 196 Cal.App.3d 223, 233 [ 242 Cal.Rptr. 37 ] [requiring general EIR discussion of contemplated pipeline].) The same principle applies here. UCSF should have discussed in the EIR at least the general effects of the reasonably foreseeable future uses of the Laurel Heights facility, the environmental effects of those uses, and the currently anticipated measures for mitigating those effects.

We do not require prophecy. The Regents are not required by our decision to commit themselves to a particular use or to predict precisely what the environmental effects, if any, of future activity will be. Nor do we require discussion in the EIR of specific future action that is merely contemplated or a gleam in a planner’s eye. To do so would be inconsistent with the rule that mere feasibility and planning studies do not require an EIR. (Guidelines, § 15262.) A detailed environmental analysis of every precise use that may conceivably occur is not necessary at this stage. (No Oil, Inc. v. City of Los Angeles, supra, 196 Cal.App.3d at pp. 235, 237-238.) The fact that precision may not be possible, however, does not mean that no analysis is required. “Drafting an EIR . . . involves some degree of forecasting. While foreseeing the unforeseeable is not possible, an agency must use its best efforts to find out and disclose all that it reasonably can.” (Guidelines, § 15144.) With the vast intellectual resources at its disposal, the University can surely make informed judgments as to probable future activities at the Laurel Heights facility.

An implicit premise of the Regents’ position is that their task will be more difficult if they must consider the environmental effects of less-than-definite future plans. This premise is flawed. We find no authority that exempts an agency from complying with the law, environmental or otherwise, merely because the agency’s task may be difficult. If CEQA is unduly burdensome, the solution lies with the Legislature, not with this court.

We hold that the EIR was inadequate because it fails to discuss the anticipated future uses of the Laurel Heights facility and the environmental effects of those uses. We cannot and do not by this opinion prescribe the exact information that the University must include in its EIR. We expect the University will attempt in good faith to fulfill its obligation under CEQA to provide sufficient meaningful information regarding the types of activity and environmental effects that are reasonably foreseeable when the remainder of the Laurel Heights facility is used by UCSF.

II.

The EIR is inadequate under CEQA because the EIR does not contain an adequate description of project alternatives.

The Association contends the EIR’s discussion of alternatives to the project is inadequate under CEQA. The Regents’ response is twofold: (1) that no discussion of alternatives was required under CEQA in light of the Regents’ conclusion that all significant environmental effects would be mitigated to a level of insignificance; and (2) that, in any event, the EIR did contain an adequate discussion of alternatives to the project. The Regents are incorrect on both counts.

A. Under CEQA, both mitigation measures and project alternatives must be discussed in an EIR.

The Regents’ argument that no discussion of alternatives was required is contrary to the express terms and underlying purposes of CEQA and to the overwhelming weight of judicial precedent. CEQA and the Guidelines are replete with references to the need for a discussion of project alternatives. Section 21002.1, subdivision (a) provides, “The purpose of an environmental impact report is to identify the significant effects of a project on the environment, to identify alternatives to the project, and to indicate the manner in which those significant effects can be mitigated or avoided.” (Italics added.) Section 21061 states that “The purpose of an environmental impact report is . . .to list ways in which the significant effects of such a project might be minimized; and to indicate alternatives to such a project.” (Italics added.) Section 21100 provides that an EIR on a project proposed by a state agency shall include a detailed statement of mitigation measures and “[alternatives to the proposed project. ” (§ 21100, subds. (c) and (d), italics added.) Perhaps most important, the Legislature has expressly declared that . . it is the policy of this state to:. . . [rjequire governmental agencies at all levels ... to consider alternatives to proposed actions affecting the environment.” (§ 21001, subd. (g), italics added.) The Guidelines require that an EIR “[djescribe a reasonable range of alternatives to the project, or to the location of the project, which could feasibly attain the basic objectives of the project and evaluate the comparative merits of the alternatives.” (Guidelines, § 15126, subd. (d).) These alternatives must be discussed, “even if these alternatives would impede to some degree the attainment of the project objectives, or would be more costly.” (Guidelines, § 15126, subd. (d)(3).)

The foregoing CEQA provisions and Guidelines make clear that “One of its [an EIR’s] major functions . . . is to ensure that all reasonable alternatives to proposed projects are thoroughly assessed by the responsible official.” (Wildlife Alive v. Chickering (1976) 18 Cal.3d 190, 197 [ 132 Cal.Rptr. 377 , 553 P.2d 537 ], italics added.)

From this statutory tapestry, the Regents extract the following thread: Section 21002 states that “. . . public agencies should not approve projects as proposed if there are feasible alternatives or feasible mitigation measures available which would substantially lessen the significant environmental effects of such projects.” (Italics added.) The Regents’ reliance on section 21002 is misplaced. It deals with public agencies’ approval of projects, not with the contents of an EIR and is thus inapposite on its face. That an agency can approve a project if environmental effects are resolved by mitigation or by alternatives does not logically mean that an EIR should not discuss both. To the contrary, requiring a discussion of both options (alternatives and mitigation measures) is consistent with CEQA’s purpose of providing responsible officials with adequate information. Indeed, the use of the word “or” in section 21002 supports the view that alternatives and mitigation measures must be discussed in an EIR because, if an agency is to assess thoroughly whether environmental effects can be alleviated by either mitigation or alternatives, the EIR must discuss both.

Even if section 21002 were meant to prescribe the matters that must be contained in an EIR, the section would not be reasonably susceptible to the interpretation urged by the Regents. Their reliance on this isolated use of the word “or” in section 21002 is contrary to the language of CEQA as a whole and to the act’s purpose. As noted above, there are numerous CEQA provisions that demonstrate a legislative intent that project alternatives must be considered. “ ‘The mere literal construction of a section in a statute ought not to prevail if it is opposed to the intention of the legislature apparent by the statute ....’” (Friends of Mammoth v. Board of Supervisors, supra, 8 Cal.3d 247, 259 , quoting In re Haines (1925) 195 Cal. 605, 613 [ 234 P. 883 ].) “We cannot, as Respondents [the Regents] would have us do, indulge in an inert exercise, leaning heavily on isolated words and phrases and remaining oblivious to the express legislative intent to protect society against environmental blight.” (Friends of Mammoth, supra, 8 Cal.3d at p. 266, fn. 9 .)

The Regents’ view also ignores the chronology of the environmental review process under CEQA. State agencies are required to certify the completion of an EIR “on any project they propose to carry out or approve.” (§ 21100.) As a matter of logic, the EIR must be prepared before the decision to approve the project. Not until project approval does the agency determine whether to impose any mitigation measures on the project. (§ 21002.1, subd. (b).) One cannot be certain until then what the exact mitigation measures will be, much less whether and to what degree they will minimize environmental effects. According to the Regents, the decision to require mitigation measures on project approval removes the need to consider project alternatives in the EIR. The decision imposing mitigation measures, however, is not made, and cannot be made under CEQA, until after the EIR has been completed. To adopt the Regents’ view, would be to say that alternatives need not be discussed if there is a possibility that the agency might adopt mitigation measures. Such result would invert the chronology of the CEQA process.

The Regents cite only one court decision in support of their position, Laurel Hills Homeowners Assn. v. City Council (1978) 83 Cal.App.3d 515 [ 147 Cal.Rptr. 842 ] (hereafter Laurel Hills). That decision does not support the Regents. In Laurel Hills, a real estate owner and developer sought approval of a 124-single-family-unit residential subdivision. The City of Los Angeles prepared an EIR that discussed the project’s environmental effects, mitigation measures, and eight alternatives to the project. The city concluded in the EIR that one of them, a 63-unit cluster alternative, was environmentally superior to the project as proposed. The city also found the proposed project would have some adverse environmental effects even after the imposition of certain mitigation measures but approved the project as mitigated despite those effects. The city, however, made no factual findings as to the feasibility of the admittedly superior cluster alternative identified in the EIR. A homeowners association challenged the approval, arguing that a finding regarding the feasibility of that alternative was mandatory under CEQA.

As is apparent from the stated facts, the question in Laurel Hills was not whether alternatives had to be discussed in the EIR. Eight alternatives were discussed, and the Laurel Hills court made clear that “It is true that an EIR must identify both feasible mitigation measures and feasible project alternatives” and referred to their “joint inclusion” in an EIR. (83 Cal.App.3d at pp. 520-521, italics added.) The court found that CEQA does not require the responsible agency to choose the environmentally best alternative identified in an EIR and stated that, if mitigation measures will avoid damage to the environment, “. . . there is no need to resort to a consideration of the feasibility of environmentally superior project alternatives identified in the environmental impact report. ” (Id., at p. 521, italics added.) The Laurel Hills court dealt only with an agency’s decision with regard to alternatives adequately discussed in the first instance in an EIR. The court found a particular decision, the choice of the best alternative, was not required. The court’s opinion, however, made clear that for the agency to make an informed decision on project approval, the EIR must meaningfully discuss alternatives. Laurel Hills was thus consistent with CEQA’s fundamental goal of fostering informed decision making. Laurel Hills does not support the Regents’ argument that only mitigation or alternatives, but not both, must be discussed in an EIR.

Finally, we note that alternatives and mitigation measures have the same function—diminishing or avoiding adverse environmental effects. The chief goal of CEQA is mitigation or avoidance of environmental harm. To argue that only mitigation measures need be discussed overlooks the fact that alternatives are a type of mitigation.

We hold that under CEQA an environmental impact report must include a meaningful discussion of both project alternatives and mitigation measures.

B. The purported discussion in the EIR of project alternatives was inadequate.

UCSF’s treatment of alternatives was cursory at best. The draft EIR identified three types of alternatives: no project anywhere, alternative sites on the UCSF Parnassus campus, and alternative sites off-campus. The three categories received a scant one and one-half pages of text in an EIR of more than 250 pages. The EIR stated the obvious conclusion that the “no project” alternative, i.e., no relocation to Laurel Heights, would not have the environmental effects identified in the EIR. It then stated in a mere two-sentence paragraph that “. . . no alternative sites on [the Parnassus] campus were evaluated as possible candidates for the location of the basic science units of the School of Pharmacy.” This is not a sufficient discussion of on-campus alternatives; it is merely an admission that such alternatives were not considered. The treatment of off-campus sites is equally deficient. It reads in its entirety: “Currently, the University has facilities at numerous other locations in the City of San Francisco, as shown in Exhibit V-l [a map]. None of these sites had space available of sufficient size to accommodate the School of Pharmacy units that are to be moved.” It defies common sense for the Regents to characterize this as a discussion of any kind; it is barely an identification of alternatives, if even that.

We agree with the Court of Appeal’s observation on this point: “Here the Regents simply referred to other facilities, designated as dots on a map of San Francisco, with no discussion of their size or available space, and with a complete lack of data to provide a factual informational underpinning for the conclusory statement that no other site had adequate space. It is impossible to analyze meaningfully the report’s conclusion that Laurel Heights is the only available facility of sufficient size. There is no assessment of the capabilities of existing sites to be expanded or remodeled with a less significant impact on the environment. There is no discussion of the possibility of purchasing or leasing other facilities, including ones not located in a high-density residential area such as Laurel Heights. . . . [H] Conclusory comments in support of environmental conclusions are generally inappropriate. (See People v. County of Kern, supra, 39 Cal.App.3d at pp. 840-842.) Moreover, the EIR’s statutory goal of public information regarding a proposed project has not been met; the EIR provides no information to the public to enable it to understand, evaluate, and respond to the bare assertion of nonavailability of alternative space. ‘The key issue is whether the selection and discussion of alternatives fosters informed decisionmaking and informed public participation.’ (Guidelines, § 15126, subd. (d)(5), italics added.)”

Even if the Regents are correct in their conclusion that there are no feasible alternatives to the Laurel Heights site, the EIR is nonetheless defective under CEQA. As we stated in a context similar to CEQA, there must be a disclosure of the “analytic route the. . . agency traveled from evidence to action.” (Topanga Assn, for a Scenic Community v. County of Los Angeles, supra, 11 Cal.3d 506, 515 [construing requirements of Gov. Code, § 65906 for zoning variances]; see also Citizens for Quality Growth v. City of Mount Shasta (1988) 198 Cal.App.3d 433, 441 [ 243 Cal.Rptr. 727 ] [construing § 21081 findings pursuant to an EIR].) The EIR prepared by UCSF contains no analysis of any alternative locations. An EIR’s discussion of alternatives must contain analysis sufficient to allow informed decision making. (San Bernardino Valley Audubon Society, Inc. v. County of San Bernardino (1984) 155 Cal.App.3d 738, 751 [ 202 Cal.Rptr. 423 ].)

The Regents argue that alternatives had already been considered and found to be infeasible during the University’s various internal planning processes and that an EIR need not discuss a clearly infeasible project alternative. The Regents apparently believe that, because they and UCSF were already fully informed as to the alleged infeasibility of alternatives, there was no need to discuss them in the EIR.

The Regents miss the critical point that the public must be equally informed. Without meaningful analysis of alternatives in the EIR, neither the courts nor the public can fulfill their proper roles in the CEQA process. We do not impugn the integrity of the Regents, but neither can we countenance a result that would require blind trust by the public, especially in light of CEQA’s fundamental goal that the public be fully informed as to the environmental consequences of action by their public officials. “To facilitate CEQA’s informational role, the EIR must contain facts and analysis, not just the agency’s bare conclusions or opinions.” (Concerned Citizens of Costa Mesa, Inc. v. 32nd Dist. Agricultural Assn. (1986) 42 Cal.3d 929, 935 [ 231 Cal.Rptr. 748 , 727 P.2d 1029 ]; People v. County of Kern, supra, 39 Cal.App.3d 830, 841-842 [conclusory statements fail to crystallize issues]; see also Citizens for Quality Growth v. City of Mount Shasta, supra, 198 Cal.App.3d 433, 441 [agency’s findings under § 21081 as to mitigation must be sufficiently detailed].) An EIR must include detail sufficient to enable those who did not participate in its preparation to understand and to consider meaningfully the issues raised by the proposed project.

If the Regents considered various alternatives and found them to be infeasible, we assume, absent evidence to the contrary, that they had good reasons for doing so. Those alternatives and the reasons they were rejected, however, must be discussed in the EIR in sufficient detail to enable meaningful participation and criticism by the public. “ ‘[Wjhatever is required to be considered in an EIR must be in that formal report; what any official might have known from other writings or oral presentations cannot supply what is lacking in the report.’ ” (Santiago County Water District v. County of Orange, supra, 118 Cal.App.3d 818, 831 [EIR found inadequate], quoting Environmental Defense Fund, Inc. v. Coastside County Water Dist. (1972) 27 Cal.App.3d 695, 706 [ 104 Cal.Rptr. 197 ].) If the Regents previously considered alternatives in their internal processes as carefully as they now claim to have done, it seems the Regents could have included that information in the EIR.

The Regents also contend the Association failed to point to any evidence in the record that demonstrates reasonable alternatives to moving the School of Pharmacy research units to Laurel Heights. This argument is somewhat disingenuous given the Regents’ own failure to provide any meaningful information regarding alternatives. It is the project proponent’s responsibility to provide an adequate discussion of alternatives. (Guidelines, § 15126, subd. (d).) That responsibility is not dependent in the first instance on a showing by the public that there are feasible alternatives. If the project proponent concludes there are no feasible alternatives, it must explain in meaningful detail in the EIR the basis for that conclusion.

The Regents’ view is also contrary to CEQA’s repeated references to the requirement that alternatives be discussed in an EIR. (See discussion at pp. 400-403, ante.) This requirement is not stated to be contingent on a showing by another party that there are feasible alternatives. Under the Regents’ view, a project proponent would never have to discuss alternatives. It would merely respond to alternatives proposed by others. There is not even a hint in CEQA that the Legislature intended such a result.

The Regents’ argument that CEQA places the burden of identifying alternatives on the Association is not supported by the cases they cite. (See, e.g., City of Lomita v. City of Torrance (1983) 148 Cal.App.3d 1062, 1069-1070 [196 Cal.Rptr 538]; Residents Ad Hoc Stadium Com. v. Board of Trustees (1979) 89 Cal.App.3d 274, 288-289 [ 152 Cal.Rptr. 585 ].) In each of those cases, the EIR contained adequate discussion of alternatives, but the party challenging the project objected by raising an additional alternative not considered in the EIR. These cases are distinguishable from the present case in which the project proponent itself failed to provide an adequate discussion of alternatives.

The Regents contend this is a distinction without a difference. We disagree. As we have explained, numerous CEQA provisions require that an EIR adequately describe feasible alternatives. Nowhere in CEQA, however, is there a provision that this duty is conditional on a project opponent coming forward with a documented alternativé. The Regents view also blinks at common sense. We can reasonably assume a project proponent knows as much or more about its project and the feasibility of various alternatives (or, conversely, why alternatives are not feasible) than anyone else, including possible project opponents. It was therefore entirely realistic that the Legislature required that alternatives be discussed in an EIR. “CEQA requires that governmental agencies consider reasonable alternatives. It is not limited to alternatives proposed and justified by objectors [to an EIR].” (Citizens of Goleta Valley v. Board of Supervisors (1988) 197 Cal.App.3d 1167, 1178 [ 243 Cal.Rptr. 339 ].)

We hold the discussion in the EIR of project alternatives is legally inadequate under CEQA. UCSF must explain in meaningful detail in a new EIR a range of alternatives to the proposed project and, if UCSF finds them to be infeasible, the reasons and facts that UCSF claims support its conclusion.

Because a new EIR is required, we believe it necessary to provide brief guidance to the parties as to the level of analysis of alternatives that must be included. As we have already explained, the analysis must be specific enough to permit informed decision making and public participation. The latter function is especially important when, as in this case, the agency approving the proposed project is also its proponent or closely related to its proponent. The need for thorough discussion and analysis is not to be construed unreasonably, however, to serve as an easy way of defeating projects. “Absolute perfection is not required; what is required is the production of information sufficient to permit a reasonable choice of alternatives so far as environmental aspects are concerned. . . . [fl] When the alternatives have been set forth in this manner, an EIR does not become vulnerable because it fails to consider in detail each and every conceivable variation of the alternatives stated.” (Residents Ad Hoc Stadium Com. v. Board of Trustees, supra, 89 Cal.App.3d at pp. 287-288; Village Laguna of Laguna Beach, Inc. v. Board of Supervisors, supra, 134 Cal.App.3d at p. 1029 .) As with the range of alternatives that must be discussed, the level of analysis is subject to a rule of reason.

III.

There is substantial evidence to support the Regents’ finding that the potential environmental effects of the project, as it is now defined in the EIR, will be mitigated.

In certifying the final EIR and approving the project, the Regents found that “changes and alterations have been required in or incorporated into the project which substantially mitigate all the significant effects as identified in the Final EIR.” This finding was required by law. (§ 21081; Guidelines, § 15091, subd. (a).) The Association contends the EIR does not provide for sufficient mitigation measures and that the Regents’ finding is thus invalid under CEQA. The central concern raised by the Association is whether there is sufficient mitigation of the potential effects on air quality and human health caused by the venting of chemicals and other substances into the air outside the Laurel Heights facility. We reject the Association’s contention that the Regents’ finding of mitigation is invalid under CEQA.

As we have explained, a court’s proper role in reviewing a challenged EIR is not to determine whether the EIR’s ultimate conclusions are correct but only whether they are supported by substantial evidence and whether the EIR is sufficient as an informational document. (See discussion at pp. 392-393, ante.) The Association, however, invites us to disregard this limitation on our review by weighing competing technical data and arguments. The Association relies on evidence in the record that the Association claims supports conclusions contrary to those reached by the Regents. The question, however, is not whether there is substantial evidence to support the Association’s position; the question is only whether there is substantial evidence to support the Regents’ conclusion.

In answering that question, the reviewing court must consider the evidence as a whole. That an EIR’s discussion of mitigation measures might be imperfect in various particulars does not necessarily mean it is inadequate. We do not suggest that a reviewing court should refrain from carefully scrutinizing the record. We have observed in a related context that such detailed review is necessary in light of the requirement that in reviewing an administrative agency’s determination the court “must scrutinize the record and determine whether substantial evidence” supports the agency’s decision. (Topanga Assn, for a Scenic Community v. County of Los Angeles, supra, 11 Cal.3d 506, 514 [zoning variances under Gov. Code § 65906].) The often technical nature of challenges to EIR’s also requires particular attention to detail by a reviewing court. The proper judicial goal, however, is not to review each item of evidence in the record with such exactitude that the court loses sight of the rule that the evidence must be considered as a whole.

Although we decline to engage in the type of balancing review sought by the Association, we will explain our conclusion that, in light of the entire record, there was substantial evidence to support the Regents’ finding of mitigation.

A. The Regents’ Evidence

We begin with the four primary types of evidence on which the Regents rely.

1. Studies at the Parnassus Campus

As the trial court properly noted, the relocation to Laurel Heights is unlike the usual CEQA case, in which the concerns are over the potential environmental effects of a new project. The Regents propose to relocate the existing Parnassus operations to another location only two miles away. The Parnassus campus research facility has been in operation for many years. The final EIR described two environmental sampling studies conducted at the Parnassus campus in 1984 and 1986, which established that research activities had not resulted in statistically significant increases in the deposition of organic chemicals or radioactive materials in the vicinity of the campus. We believe the trial court was correct in viewing these studies as evidence in support of the Regents’ finding of mitigation.

The Court of Appeal found the studies wanting in various particulars. For example, the Court of Appeal noted that these studies were of soil and vegetation and faulted them for not explaining why examination at ground

level alone is sufficient to demonstrate an absence of harmful emissions. The Regents contend such studies are more reliable and informative than air samples. The Court of Appeal also faulted the 1986 study for failing to explain that most measurements of gross gamma radiation were on the average greater than the control level. The Regents contend the court misunderstood the concept of statistical significance, which requires that a measurement exceed a control level by a given number of standard deviations before it may be considered significant. There are other disputes between the parties as to the methodology and conclusions of these studies. The Association, for example, notes that the topography of the Parnassus area differs from that at Laurel Heights.

The Court of Appeal in effect performed its own scientific critique of the studies and found the Regents should not have relied on them. This approach is inconsistent with the principle that “The court does not have the duty of passing on the validity of the conclusions expressed in the EIR, but only on the sufficiency of the report as an informative document.” (Environmental Defense Fund v. Coastside County Water Dist. (1972) 27 Cal.App.3d 695, 705 [ 104 Cal.Rptr. 197 ], italics in original.) It is also well established that “[disagreement among experts does not make an EIR inadequate.” (Karlson v. City of Camarillo (1980) 100 Cal.App.3d 789, 805 [ 161 Cal.Rptr. 260 ].)

We commend the Court of Appeal’s thoroughness in reviewing the two studies and the other evidence offered by the Regents in support of their finding of mitigation. As we have explained, such scrutiny is necessary under CEQA. The relevant point, however, is not that the two studies might be lacking in certain particulars or that the studies may not conclusively demonstrate a lack of environmental effect at the Parnassus campus or, inferentially, at Laurel Heights. Stated differently, the issue is not whether the studies are irrefutable or whether they could have been better. The relevant issue is only whether the studies are sufficiently credible to be considered as part of the total evidence that supports the Regents’ finding of mitigation. We find the studies are sufficient for that purpose. They do tend to show a lack of harmful effects at the Parnassus campus and therefore support at least an inference that the Laurel Heights operations will also have no harmful effects.

2. Absence of Evidence of Health Hazards

The Regents represented in the final EIR that they were unable to locate any evidence that the expected laboratory emissions at Laurel Heights have ever been identified as a public health hazard. The Regents explained that they had contacted numerous organizations and scientists to obtain information on emissions from fume hoods, the type of devices that would be used at Laurel Heights to vent laboratory substances into the outside air. The final EIR identified the persons contacted and stated that “The discussions with agencies and scientists did not result in any published literature on the subject of airborne emissions from fume exhaust. . . . fl|] [T]he question of emissions from a fume hood or laboratory has not been the subject of funding or study.”

The Association has failed to show that the Regents’ canvass of the scientific community was insufficient. Nor has the Association shown the Regents’ conclusion is incorrect in light of credible and significant scientific literature indicating harmful health effects of the emissions. The gist of the Association’s argument (to the extent it addresses the Regents’ showing) is that an absence of information that the emissions are harmful is not the same as information that the emissions are safe. The Association is logically correct, but the absence of scientific studies showing harm is relevant evidence. We cannot ask the Regents to mitigate the unknown.

Although the Association does not squarely contest the Regents’ representation as to an absence of scientific studies, the Association does so indirectly by citing references in the EIR to “toxic chemicals,” “air pollutants,” and “hazardous substances.” For example, the Association relies on the EIR’s statement that “Some of the chemicals that may be vented through the fume hood exhaust stacks are known to be toxic at certain levels of concentration.” That the chemicals may be toxic at certain levels and under certain circumstances, however, does not mean that the chemicals will be toxic in the levels at which they may be emitted from the fume hoods into the outside air. As noted earlier, the Regents were unable to obtain any scientific information that laboratory fume hood emissions have been shown to be harmful.

3. Absence of Regulation

The Regents also explained in the EIR that, with one exception, laboratory emissions are not subject to local, state, or federal regulations and, for the most part, are expressly exempted from such regulations. It is not clear from the record whether the Regents are entirely correct on this point. The EIR also states that the Federal Clean Air Act of 1967 established air quality standards for several pollutants, and it appears from the record that there may be other applicable regulations, for example, California Department of Health Services standards for radioactive particle concentrations in outside air. For the most part, however, the anticipated laboratory emissions are now unregulated.

The Court of Appeal rejected the Regents’ reliance on the absence of regulation, reasoning that such reliance would be similar to an illogical assertion that asbestos was not harmful before its regulation. We agree that the absence of regulation did not mean asbestos was in fact formerly safe, nor, without a showing that the responsible regulatory authorities had considered the question and affirmatively decided not to regulate, was the absence of regulation even evidence that asbestos was not harmful. But this does not mean the absence of regulation would have been entirely irrelevant to the question of the adequacy of any measures proposed to mitigate the effects of using asbestos.

The Guidelines provide that “If an emission . . . meets the existing standard for a particular pollutant, the lead agency may presume that the emission or discharge of the pollutant will not be a significant effect on the environment.” (Guidelines, § 15064, subd. (i).) This presumption does not come into play here, but the absence of any standard does indicate that the project cannnot be faulted for not limiting emissions to a prescribed level. An agency cannot be expected to predict the future course of governmental regulation or exactly what information scientific advances may ultimately reveal. “[F]oreseeing the unforeseeable is not possible.” (Guidelines, § 15144.)

We do not suggest that an absence of regulation is sufficient to show either that there will be no harmful effects or that they will be mitigated. Many harmful substances were used long before they were regulated. The absence of regulation is, however, a factor that can reasonably be considered in the EIR process. In this case, the absence of regulation was not proof of the correctness of the Regents’ finding that there would be no significant effect on air quality, but the absence of regulation was relevant to the Regents’ finding that they had taken all feasible steps to mitigate such effects in light of current scientific knowledge.

4. Regents’ Commitment to Monitor

The Regents have committed UCSF to perform baseline and periodic monitoring of ambient air quality at Laurel Heights to determine if UCSF’s research activities have any significant effect on air quality. The Attorney General, as amicus curiae, contends this commitment, although “certainly better than nothing at all,” is of questionable value absent a firm commitment by the Regents to take specific action based on the results of the monitoring. This argument misses the point. The Regents’ commitment is but one item of evidence offered in mitigation. Even if the commitment is less than overwhelming, it is entitled to consideration as part of the record as a whole. An agency’s commitment to monitor the effects of its activities may be considered as evidence of mitigation. (Twain Harte Homeowners Assn. v. County of Tuolumne (1982) 138 Cal.App.3d 664, 687-689 [ 188 Cal.Rptr. 233 ].) We also reject the argument that the commitment is not specific enough. We think it unreasonable to demand a commitment to take specific action based on unknown and as yet unknowable test results.

B. The Association’s Primary Objections

We turn now to the primary objections raised by the Association and relied on by the Court of Appeal. By noting the Association’s concerns, however, we are not engaging in the weighing process we have described as inappropriate under CEQA. As we have explained, the issue is not whether there is evidence to support the Association’s objections to the EIR, but only whether those objections show there is not substantial evidence to support the Regents’ finding of mitigation. The primary deficiencies asserted by the Association can be fairly summarized as follows:

1. Quantities of Future Emissions

The EIR does not estimate the quantities of substances that will be emitted through the fume hoods. The Association contends these substances are harmful to human health and faults UCSF for not estimating the quantities that will be involved. The Association makes no showing, however, that UCSF can reasonably do so. The changing nature of scientific research may make such estimates difficult and of little value. The proposed research laboratories are not like a manufacturing facility, whose products or history would provide a basis for reasonable estimates of future emissions. We also question the Association’s factual premise that the substances are harmful. They may be harmful in certain concentrations and under certain conditions, but, as discussed above, there is apparently no significant scientific evidence the substances are harmful as fume hood emissions.

2. Exact Substances

The Association faults the EIR for not identifying the exact nature of the chemicals to be released through the hoods. We disagree. The final EIR lists 1,200 chemicals purchased by UCSF for its School of Pharmacy research in 1985 and, for hazardous chemicals, even lists the quantities purchased. We find it entirely unreasonable to expect a scientific research facility to predict in advance every possible chemical or combination of chemicals it might conceivably use. Such result would require prescience and would shackle the scientific imagination. Again, we note that we are not faced with a manufacturing facility that could reasonably be expected to know or estimate what materials it intends to use in its operations. UCSF’s attempt to specify was sufficient.

3. Effects of Emissions

The Association contends there is a lack of factual information in the EIR regarding the effect of the anticipated emissions on human health. The gist of the Association’s argument is that the effect of emissions on human health cannot be determined because the EIR does not identify the nature and quantities of emissions. This is essentially the same argument that we have already rejected. (See p. 413, ante.) On the facts of this case, the EIR is not rendered defective merely because it does not identify every chemical that will be used and its precise quantity. We reiterate that we are dealing with an academic research facility, not an entity that could reasonably be expected to know the information sought by the Association. For example, an oil refinery or other manufacturing facility clearly would or should know in detail the nature and quantity of its emissions.

The preceding three contentions by the Association (quantities, nature, and effect of emissions) are all based on the Association’s reliance on so-called disclosures in the EIR that toxic chemicals will be used. The Association implicitly suggests there will thus be toxic emissions. Having carefully reviewed the EIR, however, we believe the Association’s suggestion is over-broad. Most of the information dealing with effects of the toxic substances on human health relates to effects on laboratory workers and those in close proximity to the chemicals. The relevant portions of the EIR are too lengthy for extensive quotation, but the following is illustrative: In section VI, subsection I, dealing with human health, the EIR states, “The potential impacts on laboratory workers due to the use of toxic materials cannot be predicted with certainty. . . . This makes it difficult, if not impossible, to set exposure standards, [fl] Harmful exposure of humans to hazardous substances may occur due to accidents, worker negligence, and unidentified risks. . . . Laboratory workers would risk exposure to hazardous materials, due to their close proximity to these chemicals. . . . Standards for workroom air, work procedures and engineering controls, as established by CAL/OSHA, are the most developed standards related to the use of toxic and radioactive compounds. . . . [fl] Toxic chemicals used in research laboratories range from low to high in mammalian toxicity. . . . Generally, mammalian toxicity is via an oral route, but can also be via inhalation and dermal routes. . . . The standards which apply to the use of these toxic chemicals are established by CAL/OSHA and apply only to workroom air, the most likely route by which laboratory workers would be effected [s/c].” (Italics added.) In short, we question the Association’s apparent assumption that a chemical that may be toxic in the closed confines of a laboratory will result in a toxic emission into the outside air.

4. Dispersion Effect

The Association contends there is inadequate information in the EIR regarding the dispersion effect of the prevailing winds. This contention is clearly untenable and reflects the unreasonable level of exactitude the Association seeks to impose on UCSF. The draft EIR explained that, “Cumulative air quality impacts would be due primarily to the use of the building for research laboratories, which would contain fume hoods for the venting and safe handling of toxic chemicals . . . .” If that were all the EIR said, we might be inclined to agree with the Association on this point, but there is much more. The draft EIR further explains that, “By law [citation], each exhaust stack would be required to be seven feet above the roof. Each would be operated by a fan. All of the fume hoods would be monitored as per standards established by [law]. . . . [fl] The Laurel Heights site is exposed to prevailing west winds, and the slope of the terrain and elevation of the site ensure good ventilation characteristics. Because the building roof is and exhaust stacks will be higher than surrounding buildings to the east, any exhausted materials would be quickly diluted and dispersed . ... [ft] While the seven foot release height of the vents and upward velocity should be sufficient for proper dilution, dilution would also occur as a result of pollutants being vented from fume hoods to the outdoor air at a rate of 1,050 cubic feet per minute.”

In response to public comment on this section of the draft EIR, UCSF provided considerable further specific information, including detailed charts in the final EIR, regarding prevailing wind characteristics and mean temperatures at the Laurel Heights facility. The Court of Appeal, however, faulted the EIR for not being specific enough, citing as examples the study evaluating the estimated efficacy of wind dispersal, the lack of discussion of wind patterns caused by surrounding buildings, and the lack of discussion of the effect of calm days. We disagree that such information was necessary: (1) A project opponent or reviewing court can always imagine some additional study or analysis that might provide helpful information. It is not for them to design the EIR. That further study of wind dispersal might be helpful does not make it necessary. (2) There is no significant evidence in the record that surrounding buildings create any significant local wind patterns. (3) The final EIR demonstrated that the air is calm only 2 percent of the time at Laurel Heights.

In short, we find the treatment of wind dispersal was adequate.

5. Absence of Regulation

The Court of Appeal found unpersuasive the Regents’ partial reliance on the near-total absence of governmental regulation of the anticipated fume-hood emissions. As we have already explained, we find such evidence to be relevant to the Regents’ finding of mitigation. (See pp. 411-412, ante.)

6. Evidence of Mitigation

The Court of Appeal found the EIR fails to present “sufficient fact-based analysis” to support its conclusion that the environmental effects will be mitigated to the point of insignificance. The Court of Appeal based its determination of this issue in large part on its finding of various inadequacies in the studies suggesting a lack of any harmful environmental effects at the Parnassus campus. For the reasons previously stated, we disagree with the conclusion that those studies were legally inadequate to support the Regents’ finding of mitigation. They were probative and relevant, if not conclusive, and were entitled to consideration as part of the evidence as a whole. (See pp. 408-409, ante.)

The Court of Appeal faulted the Regents’ conclusion in two other specific respects. First, the EIR stated that high efficiency particulate air (HEPA) filters are required for certain regulated carcinogens but did not identify these carcinogens and did not represent that the HEPA filters will in fact be used. We find these omissions insignificant. Little would be served by having UCSF promise in the EIR that it will comply with the law. Compliance can be reasonably presumed. Perhaps it would have been helpful if the specific carcinogens had been listed, but we again emphasize that the evidence must be considered as a whole.

Second, the Court of Appeal found the final EIR should have explained why all fume hood exhaust stacks will not be equipped with filters and scrubbers. The factual premise for this conclusion is a reference in the EIR to a regulatory requirement that HEPA and chemical absorbent filters be used for regulated carcinogens. We disagree that a more complete discussion of filters and scrubbers was required. The Association relies on section 15126, subdivision (c) of the Guidelines, which states, “The discussion of mitigation measures [in an EIR] shall distinguish between the measures which are proposed by project proponents to be included in the project and other measures that are not included but could reasonably be expected to reduce adverse impacts if required as conditions of approving the project.” (Italics added.) As is clear from the Guideline, only measures that can “reasonably be expected” to reduce adverse effects need be discussed. The Association’s assertion that effects on air quality “could be substantially eliminated by the installation of filters and scrubbers on all the exhaust stacks” is unsupported by the record. In the language of the Guideline, there is no showing that the use of filters and scrubbers could reasonably be expected to reduce air quality effects beyond the levels that will be achieved by the measures proposed by UCSF. For example, the majority of fume hoods on the Parnassus campus are not equipped with filters, and the two studies of emissions at that campus found no significant environmental contamination there.

C. Effects Other Than Emissions

Although the effect of emissions into the outside air is the central concern raised by the Association, it also challenges the finding of mitigation as to several other potential environmental effects of the relocation. The most significant challenges relate to traffic congestion and parking, noise, and the handling of hazardous wastes. We address these concerns in turn.

1. Traffic and Parking Effects

The draft EIR dealt at length with the potential effects of traffic and parking. The discussion of these effects occupied 42 pages of a report of 211 pages (not including notes and appendices) and included explanation of numerous traffic and parking studies. Any fair reading of the EIR demonstrates that UCSF and the Regents gave careful consideration to traffic and parking and that the effects, if any, will be mitigated. For example, the EIR cites studies projecting a 20 percent decrease in average daily traffic from the time the facility was used as a corporate headquarters to its present use by UCSF and explains that adjacent intersections will remain at “virtually identical levels of service.” The anticipated occupancy when UCSF occupies the entire building (after the CALTRANS lease expires) will be similar to that when Fireman’s Fund was present, and it appears that traffic will not appreciably increase even then as a result of UCSF’s occupancy. The EIR cited studies showing that anticipated traffic at the site and nearby intersections would be virtually the same in 1995 as when Fireman’s Fund occupied the building. Despite the apparently insignificant effects on traffic, the EIR listed several possible mitigation measures that might improve area traffic. The Association faults this discussion as being speculative. In light of the slight effect, if any, on traffic, we believe the discussion was adequate.

The discussion of parking also fails to show any significant defects. UCSF estimated that without mitigation measures 507 on-site parking spaces will be available. After mitigation measures (new spaces and restriping), there will be 547 spaces. UCSF estimated a need for 576 on-site spaces and acknowledged a deficit of 29 spaces. To eliminate this deficit, UCSF promised “to promote ongoing campus transportation systems, management programs, including promotion of transit, carpooling, vanpooling, and related activities . . . .” On these facts, we believe the discussion of mitigation was adequate.

2. Noise Effects

The EIR identified two potential effects on noise—traffic and the building’s mechanical systems, including the fume hood exhaust fans—but found there will be no increased traffic noise because there will be no increased traffic. As to mechanical noise, the EIR explained that UCSF had measured late night background sound levels at four locations around the Laurel Heights facility. The levels were found to be even lower than required by San Francisco city ordinance, but UCSF decided that such levels would be used as the performance standard for allowable noise generation at night. The daytime performance standard will be the limits allowable by city ordinance. UCSF also represented that, “The design of mechanical equipment will be reviewed by a qualified engineer prior to installation for compliance with the noise performance standards.” In response to public comments, the final EIR elaborated on mechanical noise, “The primary sources of noise anticipated from the building are the ventilation fans. The noise from these fans can be calculated once the systems are designed and the fans selected. Specific noise control treatments including fan silencers, barrier walls, or baffled enclosures will then be evaluated if predicted levels exceed the performance standards. The equipment design will be reviewed by a qualified acoustical engineer for compliance with the noise performance standards.” This is a sufficient explanation of noise mitigation measures.

3. Handling of Radioactive Substances

The Association complains of the EIR’s treatment of the handling of hazardous waste material, particularly radioactive waste. We find there was substantial evidence in the EIR that the potential effects of waste handling will be sufficiently mitigated. The draft EIR lists 14 separate mitigation measures adopted by UCSF. The discussion of radioactive waste is illustrative. It states in part, “Radioactive wastes from UCSF-Laurel Heights will be collected and disposed according to procedures described in the ‘UCSF Radiation Safety Procedure Guide.’ Anyone working with radioisotopes is required to have authorization from the Office of Environmental Health and Safety. To obtain it the applicant must pass an examination and must specify the type and amount of isotopes, in millicuries, to be used annually; the method of monitoring; the use and type of research; and must possess a certificate of competency. . . . [fl] The campus has a ‘decay program’ for isotopes with a half-life of less than 60 days. These isotopes are held in drums until they are decayed. The drums will be stored on the Laurel Heights site, in a secured location, and will be disposed as nonhazardous solid waste when sufficiently decayed. ‘Sufficiently decayed’ means to a point that radioactivity can no longer be detected, [fl] Radioactive waste with a half-life greater than 60 days is specially packaged and will be removed from the site by a licensed contractor once a week. ‘Closed-circle’ reporting is required for the disposal of this waste, which is sent from the campus to either the State of Washington or the State of Florida by the licensed contractor for final disposal. To comply with reporting requirements, UCSF must keep detailed records to assure all radioactive waste taken from the campus is received and buried at the Washington site or combusted at the Florida site.”

The Association’s primary challenge as to radioactive materials is based on prior difficulties UCSF has had in complying with regulations governing the handling of such materials. For example, in May 1986, during the EIR process, the State Department of Health Services stated in a letter to UCSF: “This letter concerns our January 14, 1986 enforcement conference relative to the radiation safety program at the Medical Center. This conference was prompted by our compliance inspection of November 6-8, 1985 which disclosed serious loss of control with respect to use of radioactive material on campus. ” (Italics added.) Having carefully reviewed the record, we agree with the Association that UCSF’s record in this regard is blemished.

Because an EIR cannot be meaningfully considered in a vacuum devoid of reality, a project proponent’s prior environmental record is properly a subject of close consideration in determining the sufficiency of the proponent’s promises in an EIR. Consideration, however, must also be given to measures the proponent proposes to take in the future, not just to the measures it took or failed to take in the past. In balancing a proponent’s prior shortcomings and its promises for future action, a court should consider relevant factors including: the length, number, and severity of prior environmental errors and the harm caused; whether the errors were intentional, negligent, or unavoidable; whether the proponent’s environmental record has improved or declined; whether he has attempted in good faith to correct prior problems; and whether the proposed activity will be regulated and monitored by a public entity. In this case, these factors weigh in favor of UCSF: (1) There is no evidence UCSF’s compliance difficulties resulted in any severe danger or adversely affected human health in the slightest degree. (2) There is no evidence of intentional violation. (3) UCSF appears to have attempted in good faith to remedy its problems with radioactive substances. (4) The handling of radioactive substances is closely regulated and monitored, as evidenced by the oversight of UCSF’s activities.

The present EIR can be fairly read as a firm commitment by UCSF to comply with sound practices, which are detailed in the EIR. In consideration of that commitment and the other factors discussed above, we conclude that, on balance, there is substantial evidence to support the finding of mitigation of handling of radioactive wastes at Laurel Heights.

The Association also contends that, according to state inspectors and UCSF documents, significant amounts of radioactive particles are emitted into the air during laboratory operations. This assertion is a gross misstatement of the record. In March 1985, the State Department of Health Services inquired by letter as to a provision of UCSF’s safety manual that prohibits radiation discharge via the ventilation exhaust system and asked how this would be accomplished “during laboratory operations where a significant portion of the radioactive material may become airborne.” (Ital ics added.) UCSF responded in April 1985 as follows: “Airborne isotopes, i.e. gases, other than xenon-133 are not normally used on campus. The institution does not plan to willfully exhaust radioisotopes up the hoods, other than the previously mentioned xenon-133. Those isotopes which may become volatile are to be used in the hoods. Currently most hoods on the [Parnassus] campus do not contain HEP A filters, thus resulting in possible loss of volatile isotopes to the environment. We are establishing an environmental monitoring program for assessment of the amount of contamination that may be occurring, if any, about campus and on the buildings from the hood exhausts.” UCSF’s response demonstrates two things. First, it was not a disclosure, as the Association contends, that “significant” amounts of radiation are being emitted into the outside air. The response indicates only that some emissions may occur, and even that appears to be doubtful. Second, the response demonstrated a commitment to monitor.

Such monitoring did in fact take place and it suggests the Association’s fears as well as its assertions regarding radioactivity are greatly exaggerated. The final EIR states, “To ascertain the potential for any radioactive materials to enter the atmosphere and be deposited around the campus, the UCSF Office of Environmental Health and Safety in June 1986, conducted an analysis of seven locations upwind and downwind of the [Parnassus] campus to survey for gross gamma, hydrogen-3 and carbon-14. . . . []f] The University concluded from these tests that there is little if any discernible difference between testing for radioactive materials or chemical materials resulting from fume hood exhaust at the UC San Francisco Parnassus campus when measured upwind or downwind of the campus.”

It also appears from the record that very low levels of radiation are used in the laboratories that will be transferred to Laurel Heights. For example, the Inter-Campus Program, which the Association challenges in particular, uses only minute quantities of radioactive substances, measured in microcuries (millionths of a curie) per use, or millicuries (thousandths of a curie) per year. By contrast, many major hospitals, such as Moffett Hospital on the UCSF campus, are licensed to use as many as tens of curies per year of the same substances—thousands of times the amount used in Inter-Campus Program research.

In light of the evidence we have discussed and the record as a whole, we find substantial evidence to support the Regents’ finding of mitigation as to the use of radioactive substances.

D. Conclusions as to Mitigation Finding

Summarizing our consideration of the Regents’ finding of mitigation, we conclude that, when considered individually, some of the evidence proffered by the Regents is subject to reasonable criticism. The EIR might be a more useful document if it included some of the information the Association seeks. That some items or types of evidence, however, are less than conclusive does not mean the evidence as a whole is not substantial. As we have explained in a related context, “the reviewing court must resolve reasonable doubts in favor of the administrative findings and decision.” (Topanga Assn, for a Scenic Community v. County of Los Angeles, supra, 11 Cal.3d at p. 514 .) To paraphrase the Guidelines, a fair argument can be made to support the Regents’ conclusion, even though other conclusions might also be reached. (Guidelines, § 15384, subd. (a).) We hold that, taken as a whole, there was substantial evidence to support the Regents’ conclusion that the environmental effects of the project, as now defined, will be mitigated.

Our holding as to mitigation does not apply to additional environmental effects that may be identified when a new EIR is prepared in accordance with our holdings that the project is not properly defined and that alternatives are not adequately discussed. To the extent that different or additional effects (or the magnitude of such effects) are identified when the project is properly defined in the new EIR, the Regents must comply with CEQA by sufficiently identifying in the new EIR measures to mitigate those effects and by making the required finding or findings as to mitigation. (§21081; Guidelines, § 15091, subd. (a).) The mitigation discussion in the present EIR of effects already identified is sufficient, however, and need not be revised in the new EIR.

IV.

UCSF may continue its present activities at Laurel Heights pending certification of an adequate EIR.

Our decision that the EIR is inadequate under CEQA raises the question of whether UCSF should be allowed to continue its present activities at Laurel Heights pending the Regents’ certification of a proper EIR. We conclude that such activities may continue.

As originally enacted, CEQA did not prescribe the action that a court should take if it found that CEQA had been violated. In 1984, section 21168.9 was enacted. We must construe this statute for the first time. Subdivision (a) requires that if a court finds “any determination, finding, or decision of a public agency has been made without compliance with this division” the court shall enter an order with one or more specified provisions. One of them is “[a] mandate that the public agency and any real parties in interest suspend all activity, pursuant to the finding, determination, or decision, that could result in any change or alteration to the physical environment, until the public agency has taken such actions as may be necessary to bring the determination, finding, or decision into compliance with this division.” (§ 21168.9, subd. (a)(2).) Section 21168.9 grants us the authority to stay all activity at Laurel Heights until the Regents certify a proper EIR. The question is whether we should do so.

Because CEQA does not require us to enjoin the present activity, we rely on traditional equitable principles in deciding whether injunctive relief is appropriate. We find guidance in recent decisions of the United States Supreme Court in cases arising under federal environmental statutes. In Weinberger v. Romero-Barcelo (1982) 456 U.S. 305 [ 72 L.Ed.2d 91 , 102 S.Ct. 1798 ], the United States Navy had been found by a federal district court to be in violation of the Federal Water Pollution Control Act (FWPCA) because the Navy had not obtained a National Pollutant Discharge Elimination System permit for weapons training during which naval ordnance fell into the sea off the Puerto Rico coast. The district court ordered the Navy to apply for the required permit but allowed the Navy to continue its activities while seeking the permit. The Court of Appeals found that FWPCA withdrew the courts’ equitable discretion and required an injunction of all activities until the Navy obtained the permit. The high court disagreed and reasoned that traditional equitable principles govern the decision of whether to enjoin activity that violates the FWPCA.

The court rejected the argument that failure to enjoin the Navy would undermine the integrity of the permit process by allowing the statutory violation to continue. The court explained that, “The integrity of the Nation’s waters . . . not the permit process, is the purpose of the FWPCA.” ( 456 U.S. at p. 314 [ 72 L.Ed.2d at p. 100 , 102 S.Ct. at p. 1804 ].) The court recognized that compliance with the permit requirements achieves the purpose of protecting water resources but explained that such purpose would not be undercut by allowing the Navy to continue because it had been ordered to obtain the permit and because the activity had not polluted the water. ( 456 U.S. at p. 315 [ 72 L.Ed.2d at p. 100 , 102 S.Ct. pp. 1804-1805].)

We agree with the high court’s approach and adopt it for deciding whether a CEQA violation must be enjoined. A primary purpose of CEQA is to protect the environment. In light of our conclusion that there is substantial evidence to support the Regents’ finding that the present activities will be mitigated, we believe CEQA will not be thwarted by allowing UCSF to continue its present activities at Laurel Heights. We believe the Regents have the good faith and ability to prepare an EIR that complies with CEQA and that they will proceed apace to do so. We can reasonably assume the Association and the trial court will closely monitor the Regents’ progress in complying with our decision. Such oversight is an additional assurance a new EIR will be completed without undue delay. Should it become clear, however, that the Regents cannot or will not prepare and certify a legally adequate EIR and that compliance with CEQA will not be promptly forthcoming, the trial court can reconsider the question of whether equitable relief terminating operations at Laurel Heights is then appropriate.

As in Romero-Barcelo , there is no evidence the environment is being adversely affected by the present activities. We have expressly found there is substantial evidence of mitigation of the potential effects of the present activities identified in the present EIR. The defects in the EIR relate only to future activity, which the EIR failed to address, and to feasible alternatives.

UCSF and the general public might be unduly prejudiced if we were to enjoin the present activities. Requiring the University to cease existing laboratory operations at the Laurel Heights facility and to move them to other sites would cause unnecessary cost that would ultimately be borne by the taxpayers. More important, such an order would seriously disrupt ongoing scientific research and perhaps cause the University to lose important faculty members and research funds. UCSF’s research is designed to improve the state of medical knowledge and thus improve and even save lives. We are especially reluctant to interfere unnecessarily with such a salutary enterprise.

Having duly considered the equities, we hold that UCSF may continue operations that have already begun at Laurel Heights as of the date this opinion is filed but that UCSF may not expand existing operations at Laurel Heights or begin additional operations there, whether or not identified in the present EIR, until a new EIR is certified and the project reapproved by the Regents. This restriction on expansion will provide the Regents with an incentive to act promptly to remedy the defects in the present EIR. Having failed to comply with CEQA in the first instance, they cannot fairly complain of any burden that preparing a new EIR or a restriction on expansion might impose on them.

One additional point merits brief discussion to guide the Regents in their future efforts to comply with CEQA. We emphasize that neither the present activity we are allowing to continue nor any prior UCSF activities involving Laurel Heights (including planning for the relocation) can serve as a proper basis for rejecting feasible alternatives to the Laurel Heights site. We shall not countenance any attempt to reject an alternative on the ground that the Laurel Heights site has already been purchased or that activities there have already commenced. The Regents must begin anew the analytical process required under CEQA. We will not accept post hoc rationalizations for actions already taken, particularly in light of the fact that those activities were begun in violation of CEQA, even if done so in good faith. To do so would tarnish the integrity of the decisionmaking process required by CEQA. Stated differently, the Regents have begun the relocation despite their failure to comply with CEQA in the first instance. It would be untenable for them to rely on the result of their own noncompliance as a basis for determining their future action.

V.

The Court of Appeal was not required to remand to the trial court for initial determination the question of whether the Association is entitled to attorneys fees under Code of Civil Procedure section 1021.5.

After issuing its decision on the merits, the Court of Appeal granted the Association’s motion for an award of attorneys fees under Code of Civil Procedure section 1021.5 and remanded the matter to the trial court for the limited purpose of determining the amount of the award. The Regents contend the Court of Appeal should have remanded the action to the trial court to determine in the first instance whether the Association is entitled to a fee award under section 1021.5. We disagree.

The Regents do not contend that the Court of Appeal lacked jurisdiction or that it abused its discretion. They argue only that the normal procedure when a judgment is reversed on appeal is for the appellate court to remand the fee question under section 1021.5 to the trial court. We agree that in many, perhaps most, cases, including those under CEQA, the trial court will be better equipped to decide whether fees should be awarded under section 1021.5. It is therefore proper for a reviewing court to defer to the trial court in making that determination. (Woodland Hills Residents Assn., Inc. v. City Council (1979) 23 Cal.3d 917, 940-941 [ 154 Cal.Rptr. 503 , 593 P.2d 200 ]; Citizens for Quality Growth v. City of Mount Shasta, supra, 198 Cal.App.3d at p. 448 ; Citizens Assn, for Sensible Development of Bishop Area v. County of Inyo, supra, 172 Cal.App.3d at pp. 176-177.) That a procedure is normal or customary, however, does not make it mandatory. Section 1021.5 does not require the initial determination to be made by the trial court.

Our recent decision in Harbor v. Deukmejian (1987) 43 Cal.3d 1078 [ 240 Cal.Rptr. 569 , 742 P.2d 1290 ] is illustrative. A petition for a writ of mandate was originally filed in the Court of Appeal and denied by that court. We granted review and affirmed the denial. For present purposes, our functional role in that case was analogous to that of the Court of Appeal, whose role was analogous to that of a trial court. We granted the petitioners’ request for attorneys fees under section 1021.5 and remanded to the Court of Appeal to determine the amount of fees.

The Regents’ reliance on our decision in Woodland Hills Residents Assn., Inc. v. City Council, supra, 23 Cal.3d 917 is misplaced. The Regents correctly point out that we did remand to the trial court for the initial determination as to whether fees should be awarded, but the Regents overlook the key fact that the trial court had denied the plaintiff’s motion for fees before the effective date of section 1021.5. We held the then newly enacted statute applied to the plaintiff’s action and remanded for the trial court to reconsider its decision in light of the statute. Because the trial court was already familiar with the fees issue, it made practical sense for that court to determine it in the first instance. (See also Baggett v. Gates (1982) 32 Cal.3d 128, 143 [ 185 Cal.Rptr. 232 , 649 P.2d 874 ] [reversal of trial court’s denial of fees under section 1021.5].) In the present case, the issue of attorneys fees was fully briefed to the Court of Appeal, and it was as capable as the trial court of deciding the issue. More important, we reiterate that, even though the better practice may be to remand to the trial court for the fees determination, section 1021.5 does not require a remand for that purpose.

We agree with the reasoning in Wilkerson v. City of Placentia (1981) 118 Cal.App.3d 435 [ 173 Cal.Rptr. 294 ], in which the Court of Appeal awarded fees under section 1021.5 and remanded to the trial court only for a determination as to the amount of fees. The court appropriately noted that “where the entire basis for the decision is what we have done in this opinion, we should make the decision that it qualifies as a case for granting of fees.” (Id., at p. 445.) Likewise here, it was the Court of Appeal’s decision, not the trial court’s, which provided a basis for fees in the first instance.

We hold that a Court of Appeal is not required to remand to the trial court the question of whether to award fees under section 1021.5. The decision as to which court is better equipped to make the initial determination is a matter within the discretion of the Court of Appeal. We find no abuse of that discretion in this case.

Disposition

For the reasons discussed above, we affirm the Court of Appeal’s decision that the EIR should have addressed anticipated future uses and their environmental effects and that the discussion of project alternatives is inadequate under CEQA. We also affirm the Court of Appeal’s decision that the Association is entitled to its attorneys fees under section 1021.5. We reverse the Court of Appeal’s decision that the Regents’ finding as to mitigation is inadequate.

The cause is remanded to the Court of Appeal with directions to order the trial court: (1) to grant the Association’s petition for a writ of mandate vacating the Regents’ certification of the EIR and approval of the project, and (2) to determine promptly the amount of attorneys fees that should be awarded to the Association for its efforts in this action. Payment of such award shall not be delayed pending final outcome of this action.

Pending the Regents’ certification of a proper EIR and reapproval of the project, the Court of Appeal is directed to allow UCSF to continue operations that have already begun at Laurel Heights as of the date this opinion is filed to the extent those activities are described in the present EIR, but not to allow UCSF to expand existing operations at Laurel Heights or begin additional operations there until a proper EIR is certified.

Our order dated August 12, 1987, prohibiting use of radioactive materials is modified as set forth in footnote 25 of this opinion. (See p. 425, ante.)

Pursuant to Public Resources Code section 21168.9, subdivision (b), the Court of Appeal shall order the trial court to retain jurisdiction over this action and to specify promptly, after notice and hearing, a date by which the Regents must certify a new EIR in accordance with CEQA standards and procedures, including provisions for public comment, and to make any findings that may be required by CEQA.

Lucas, C. J., Mosk, J., Broussard, J., Panelli, J., Arguelles, J., and Kaufman, J., concurred.

Appellant’s petition for rehearing was denied January 26, 1989, and the opinion was modified to read as printed above.

Except in part V of this opinion, all statutory references are to the Public Resources Code unless otherwise indicated. All references to “Guidelines” are to the State CEQA Guidelines, which implement the provisions of CEQA. (Cal. Code Regs., tit. 14, § 15000 et seq.) The Guidelines state, “These Guidelines are binding on all public agencies in California.” (Guidelines, § 15000.) This court, however, has not decided the issue of whether the Guidelines are regulatory mandates or only aids to interpreting CEQA, and the Courts of Appeal have differed somewhat on this issue. (City of Santa Ana v. City of Garden Grove (1979) 100 Cal.App.3d 521, 528-529 [ 160 Cal.Rptr. 907 ] [suggesting binding effect of Guidelines]; Karlson v. City of Camarillo (1980) 100 Cal.App.3d 789, 804-805 [ 161 Cal.Rptr. 260 ] [suggesting nonbinding effect]; see generally Selmi, The Judicial Development of the California Environmental Quality Act (1984) 18 U.C. Davis L.Rev. 197, 280-284 [discussing differences among Court of Appeal interpretations].) Whether the Guidelines are binding regulations is not an issue in this case, and we therefore need not and do not decide that question. At a minimum, however, courts should afford great weight to the Guidelines except when a provision is clearly unauthorized or erroneous under CEQA. (Rural Landowners Assn. v. City Council (1983) 143 Cal.App.3d 1013, 1022 [ 192 Cal.Rptr. 325 ].) The Regents do not dispute that the University of California is a “public agency” (§ 21063) and that the relocation to Laurel Heights is a “project” under CEQA (§ 21065). “ ‘Lead agency’ means the public agency which has the principal responsibility for carrying out or approving a project which may have a significant effect upon the environment.” (§ 21067.) The parties dispute whether the Association’s challenge to the Regents’ certification of the EIR and approval of the project was a traditional or administrative mandamus proceeding. Section 21168 establishes the standard of review in administrative mandamus cases. Section 21168.5, quoted above, governs traditional mandamus actions. This action appears to be one of traditional mandamus because the agency did not conduct a hearing at which evidence was taken in a judicial (adjudicative) sense, but we need not decide this issue. As the Court of Appeal noted, the dispute is mostly academic because the standard of review is essentially the same under either section, i.e., whether substantial evidence supports the agency’s determination. (B owman v. City of Petaluma (1986) 185 Cal.App.3d 1065, 1072, fn. 7 [ 230 Cal.Rptr. 413 ]; Citizens Assn, for Sensible Development of Bishop Area v. County of Inyo (1985) 172 Cal.App.3d 151, 164 [ 217 Cal.Rptr. 893 ].) The Court of Appeal agreed with the Association, finding that, “The EIR should have addressed the issue of the extent and cumulative impact of the anticipated future plans,” and that the failure to do so rendered the EIR’s “project description” legally inadequate under CEQA. Although we agree with the result reached by the Court of Appeal on this issue, we believe the issue should not be rigidly defined as whether the project description was adequate. As we understand the parties’ arguments, the fundamental dispute is whether the EIR adequately discussed future uses of the Laurel Heights facility and their environmental effects. Thus, the challenge is not just to the project description but to the EIR’s discussion of future environmental effects. In light of this holding, we need not and do not respond to the Regents’ argument that only they, rather than subordinate officials, have the authority to approve UCSF’s relocation to Laurel Heights and the uses of that facility. With a project of this magnitude, the Guidelines appear to provide for precisely the present situation. Section 15168, subdivision (a) explains that an agency can use a “program EIR,” for “a series of actions that can be characterized as one large project and are related either: (1) Geographically, (2) As logical parts in the chain of contemplated actions,... or (4) As individual activities carried out under the same authorizing statutory or regulatory authority and having generally similar environmental effects which can be mitigated in similar ways.” Section 15385 of the Guidelines provides for “tiering” of EIR’s, which is “the coverage of general matters in broader EIRs (such as on general plans or policy statements) with subsequent narrower EIRs . . . .” The Laurel Heights relocation facility appears to be an excellent candidate for these types of EIR’s. The Regents’ argument that alternatives need not be discussed in an EIR rings hollow in light of the fact that the Regents discussed them in the present EIR, albeit insufficiently. We doubt the Regents would have done so unless they thought it necessary. “ ‘Feasible’ means capable of being accomplished in a successful manner within a reasonable period of time, taking into account economic, environmental, legal, social, and technological factors.” (Guidelines, § 15364.) Section 21081 states in part: “[N]o public agency shall approve or carry out a project for which an environmental impact report has been completed which identifies one or more significant effects thereof unless such public agency makes one, or more, of the following findings: [H] (a) Changes or alterations have been required in, or incorporated into, such project which mitigate or avoid the significant environmental effects thereof as identified in the completed environmental impact report.” We do not suggest that a court must uncritically rely on every study or analysis presented by a project proponent in support of its position. A clearly inadequate or unsupported study is entitled to no judicial deference. The Association, however, has failed to demonstrate that the two studies are clearly inadequate or unsupported. As is apparent from the foregoing discussion (see pp. 408-410, ante), the studies at the Parnassus campus and the absence of evidence of health hazards might logically tend to show an absence of certain effects in the first instance rather than mitigation of those effects. The Regents, however, have relied on these types of evidence to show mitigation rather than an absence of effects, and the parties have framed the issue as being whether there is sufficient evidence to support the finding of mitigation. We have discussed the issue as framed by the parties. The EIR stated, “The Bay Area Air Quality Management District currently enforces emissions limitations for five toxic air contaminants, but does not specify outdoor concentration. The five materials are asbestos, beryllium, lead, mercury, and vinyl chloride. None of these are used in the pharmacy research activities, [fl] In December 1982, the California Air Resources Board (CARB) began to develop policies related to the control and study of toxic air pollutants. CARB is studying non-regulated toxic emissions to the air basin which could result in standards and controls that affect UCSF. [fl] No standards are established either by OSHA [Occupational Safety and Health Administration], the Bay Area Air Quality Management District (BAAQMD), or the California Air Resources Board (CARB) for emissions from fume hoods except for requiring High Efficiency Particulate Air (HEPA) filters and chemical absorbent filters for regulated carcinogens.” Apparently in acknowledgment of scientific advances, the EIR states, “In the future, research and ultimately environmental standards and controls will extend to numerous other substances, and may apply to chemicals that are not now controlled.” We also note that the Regents submitted the preliminary results of a new study of air emissions at Laurel Heights in opposition to the Association’s motion to the Court of Appeal for a stay. The Association faults the study in several particulars, but we think it worthy of mention. It was conducted by the Radian Corporation, an independent environmental consultant. The study found that, even using conservative factual assumptions, the estimated human-exposure level to the substances tested fell below the safe-exposure level in every case by a factor of at least 500 and, in some cases by a factor of as much as 500 million. This study was not part of the EIR process, and we therefore do not rely on it as the basis for any conclusion. We merely note the study because it tends to confirm the Regents’ commitment to monitor and also puts into perspective the Association’s dire predictions as to emissions. The Association also objects to the fact that this list was included in the final EIR after the public comment period had expired. Although earlier inclusion of this list would have been preferable, we find the slight delay of little consequence, especially in light of our holding that a new EIR must be prepared. The list can be included in the new EIR and will thus be available for public comment at that time. The draft EIR explained that “any level of exposure” to radioactive substances “may pose a potential hazard,” and that “Some of the chemicals that may be vented through the fume hood exhaust stacks are known to be toxic at certain levels of concentration.” Some of the chemicals identified in the list of 1,200 purchased by the School of Pharmacy are described in Environmental Protection Agency chemical profiles as highly toxic in small amounts. Although the Court of Appeal centered its discussion on the effects of emissions, it directed the Regents to pay careful attention to these other potential effects in preparing a new EIR. On August 6, 1987, we stayed enforcement of the Court of Appeal’s order staying all activity by UCSF at Laurel Heights. We subsequently modified our order to reinstate the Court of Appeal’s order to the extent that it stayed the introduction of radioactive materials to Laurel Heights. According to the EIR, the California Department of Health Services regulates and monitors the use of radioactive substances, and various other governmental agencies have established guidelines for handling hazardous wastes. In response to public questions on the draft EIR, the final EIR stated, “A California State Department of Health permit is required for disposal of these [carcinogenic] hazardous wastes. . . . [fl] The University of California Procedures follow guidelines issued by the Occupational Safety and Health Administration (OSHA), the Environmental Protection Agency (EPA), the Department of Transportation (DOT), and the State Department of Health as well as the San Francisco Department of Health in its handling of hazardous materials.” The Association has challenged the Regents’ finding of mitigation on other, apparently less significant, grounds in addition to those discussed in this opinion. We find it unnecessary to discuss those additional objections, but we have carefully reviewed the record, and our holding as to mitigation applies to all objections raised by the Association. The high court recently affirmed the importance of traditional equitable principles in deciding whether to enjoin activity that violates an environmental statute. (Amoco Production Co. v. Village of Gambell, Alaska (1987) 480 U.S. 531 [ 94 L.Ed.2d 542, 553-556 , 107 S.Ct. 1396, 1402-1405 ] [construing the Alaska National Interest Lands Conservation Act].) Tradi tional equitable principles also govern in actions under the National Environmental Policy Act. (State of Wis. v. Weinberger (7th Cir. 1984) 745 F.2d 412, 424-428 .) Section 21168.9, subdivision (b), provides that the trial court shall retain continuing jurisdiction over the agency’s proceedings until the court determines that the agency has complied with CEQA. In light of our holding preventing the commencement of new activities pending certification of a proper EIR and our finding that there is substantial evidence of adequate mitigation measures for radioactive materials (see pp. 419-421, ante), our order dated August 12, 1987, staying the introduction of radioactive materials to the Laurel Heights facility is modified as follows: Pending certification of a new EIR and reapproval of the project by the Regents, radioactive materials may be introduced at the Laurel Heights facility only to the extent they are used in direct connection with those activities already begun at Laurel Heights and allowed to continue under this opinion. This modification does not exempt UCSF from obtaining all required licenses and permits for the use of radioactive materials before introducing them to the Laurel Heights facility. Our order, as modified, shall expire when a new EIR is certified and the project reapproved by the Regents. Section 1021.5 provides in part. “Upon motion a court may award attorneys’ fees to a successful party against one or more opposing parties in any action which has resulted in the enforcement of an important right affecting the public interest if: (a) a significant benefit, either pecuniary or nonpecuniary, has been conferred on the general public or a large class of persons, (b) the necessity and financial burden of private enforcement are such as to make the award appropriate, and (c) such fees should not in the interest of justice be paid out of the recovery, if any.” In this part V of our opinion, all statutory references are to the Code of Civil Procedure unless indicated otherwise. To the extent they are contrary to our decision in this case, we disapprove of the decisions in Los Angeles Police Protective League v. City of Los Angeles (1985) 163 Cal.App.3d 1141, 1149-1150 [ 209 Cal.Rptr. 890 ] and Los Angeles Police Protective League v. City of Los Angeles (1985) 166 Cal.App.3d 55, 65-66 [ 212 Cal.Rptr. 251 ], in which the court expressly found the trial court should decide whether to award fees and disapproved of the decision in Wilkerson v. City of Placentia, supra, 118 Cal.App.3d 435 . As explained above, we find Wilkerson to state the correct view. In determining the amount of fees, the trial court should consider the fact that the Association was ultimately unsuccessful in its challenge to the Regents’ finding of mitigation, discussed in part IV of this opinion.