INTRODUCTION

In this consolidated case, the International Brotherhood of Teamsters, Amalgamated Transit Union, Railway Labor Executives' Association, and Oil, Chemical, and Atomic Workers Union (collectively, “the Unions”) petition for review of an order issued by the Federal Highway Administration of the Department of Transportation (“FHWA”) mandating various forms of drug testing for commercial motor vehicle operators. The Unions argue that the planned implementation of random, pre-em-ployment, post-accident, and biennial drug testing violates the drivers’ fourth amendment right against unreasonable searches. The Unions also contend that the FHWA acted arbitrarily and capriciously in promulgating the regulations. We hold that these regulations do not run afoul of the fourth amendment. Neither was the FHWA’s decision to promulgate the regulations arbitrary or capricious. Accordingly, we deny the petition for review.

BACKGROUND

A. The Testing Program

On June 14, 1988, the FHWA published a notice of proposed rulemaking, in which it proposed that certain commercial drivers be tested for the use of controlled substances. 53 Fed.Reg. 22,268 (1988). Over the ensuing months, the FHWA received approximately 145 comments concerning the proposed drug testing program. 53 Fed.Reg. 47,134, 47,136 (1988). Additionally, 43 persons and organizations testified at four public hearings conducted by the FHWA. Id.

On November 21, 1988, the FHWA published its final drug testing regulations. 53 Fed.Reg. at 47,151 (1988) (codified at 49 C.F.R. §§ 391.81-391.123 (1989)). The announced purpose of the testing program was “[t]o detect and deter the use of drugs by bus and truck drivers.” 53 Fed.Reg. at 47,135. The FHWA reasoned that:

Id.

While recognizing drivers’ privacy concerns about the program, the FHWA con- *1291 eluded that “the clear public interest in assuring that commercial motor vehicle drivers perform their duties free of prohibited substances” outweighed the intrusion on drivers’ expectations of privacy. Id. at 47,137 . The FHWA conceded from the outset that data documenting actual drug use by commercial drivers and its role in highway accidents were scarce. 53 Fed.Reg. 22,268, 22,270. The FHWA attributed this dearth of evidence, at least in part, to the fact that commercial drivers generally work without supervision and that epidemiological studies of accident causes were in their nascency. 53 Fed.Reg. at 22,269. Hearings on the proposed regulations produced testimonial evidence and empirical studies corroborating the existence of a drug problem among commercial drivers, but left undetermined the extent of the problem. The FHWA felt, nonetheless, that the combination of evidence adduced at the hearings and the pervasiveness of drug abuse in society in general justified the promulgation of the drug testing regulations. 53 Fed.Reg. at 47,137.

The regulations require all interstate motor carriers to implement drug testing programs for drivers operating vehicles (i) weighing more than 26,000 pounds, (ii) carrying fifteen or more passengers, or (iii) transporting hazardous materials. 49 C.F.R. § 391.85 . The carriers must test all employee drivers and all contract drivers who are under contract with them for ninety or more days a year. The FHWA estimated that the regulations would affect 200,000 motor carriers and three million drivers. 53 Fed.Reg. 47,149.

The regulations prescribe the institution of drug tests in six different instances. First, drivers must submit to random testing. 49 C.F.R. § 391.109 . The random selection of drivers to be tested must be done “in a scientifically acceptable manner.” 53 Fed.Reg. 47,134, 47,137. Random tests must be conducted at a rate at least equal to fifty percent of the eligible drivers per year. 49 C.F.R. § 391.109 . 1

Second, the regulations require biennial testing of drivers. 49 C.F.R. § 391.105 . Carriers are instructed to conduct the first test in conjunction with the routine medical examinations of their drivers. 49 C.F.R. § 391.105 (a). The regulations permit employers to cease biennial testing once they have fully implemented their random drug testing programs. 49 C.F.R. § 391.105 (c).

Third, motor carriers must conduct pre-employment drug tests. 49 C.F.R. § 391.103 . Testing is limited to driver-applicants whom the carrier intends to hire. 49 C.F.R. § 391.103 (a).

Fourth, drivers involved in “reportable accidents]” must arrange to take a drug test within thirty-two hours of the incident. 49 C.F.R. § 391.113 . The regulations consider an accident “reportable” if it involves (1) a fatality, (2) an injury demanding immediate medical treatment away from the scene of the accident, or (3) at least $4,400 in property damage. 49 C.F.R. § 394.3 .

Fifth, the regulations require testing when the employer has cause to believe that a driver is using a controlled substance. 49 C.F.R. § 391.99 . The conduct inspiring the reasonable suspicion must have been witnessed by at least one supervisor trained in the detection of probable drug use. 49 C.F.R. § 391.99 (c).

Sixth, drivers who test positive for drug use must submit to follow-up testing for up to five years after their return to work. 49 C.F.R. § 391.123 .

The Unions do not challenge the propriety of the latter two drug testing programs (reasonable suspicion and follow-up). They confine their petitions to attacking the validity of the random, biennial, pre-employment, and post-accident testing schemes.

B. Drug Testing Procedures

The regulations incorporate the Department of Transportation’s procedures for conducting drug tests. 49 C.F.R. §§ 391 .-81(b)-391.123 (incorporating in various sections the procedures set out at 49 C.F.R. §§ 40.1-40.41 ). As a result, the drug test *1292 ing collection site must be properly secured and must provide a private enclosure within which urination can occur. Direct observation of urination is allowed only in narrowly defined circumstances where the monitor may reasonably suspect the integrity of the specimen. 2 49 C.F.R. § 40.25 (e). The collection procedure must be supervised by either a licensed medical professional or an individual specifically trained in specimen collection procedures. 49 C.F.R. § 40.23 (d). Strict collection and chain of custody procedures must be followed. 49 C.F.R. §§ 40.23 , 40.25(f).

The regulations likewise prescribe the qualifications and testing procedures to be followed by laboratory personnel in analyzing the specimens. 49 C.F.R. §§ 40.27 , 40.-29, 40.31, 40.33. Employers may only use laboratories certified under the Department of Health and Human Services’ Mandatory Guidelines for Federal Workplace Drug Testing Programs. 49 C.F.R. § 40.39 .

The laboratories initially perform an immunoassay test on the specimen. 49 C.F.R. § 40.29 (e). All positive readings must then be confirmed through use of gas chromatography/mass spectrometry techniques. 49 C.F.R. § 40.29 (f).

Positive results of drug tests are reported first to the employer’s designated Medical Review Officer (“MRO”). 49 C.F.R. § 40.33 (a). The MRO must be a licensed physician with knowledge of substance abuse disorders. 49 C.F.R. § 40.33 (b). Prior to disclosing the result to the employer, the MRO must investigate alternative explanations for the positive result and must provide the employee an opportunity to explain the result. 49 C.F.R. § 40.33 (b) & (c). If the positive reading is attributable to legitimate medical treatment, the MRO shall report the test result to the employer as negative. 49 C.F.R. § 40.33 (f). Similarly, if the MRO determines the result is scientifically insufficient or unreliable, the test result shall be reported as negative. 49 C.F.R. § 40.33 (g).

The regulations oblige laboratories to treat employee test records confidentially. 49 C.F.R. § 40.35 . The MRO may not release test results to anyone, other than the employer, absent the written consent of the employee. 49 C.F.R. § 391.89 .

Finally, each employer must establish, in conjunction with the implementation of drug testing, an Employee Assistance Program (“EAP”). 49 C.F.R. § 391.119 . The EAPs are designed to promote safety and reduce drug use in the motor carrier industry. 53 Fed.Reg. 47,134, 47,154 (1988). The EAP must, at a minimum, include educational information and training for drivers and supervisory personnel about controlled substances. 49 C.F.R. § 391.119 . Such information must include advice about the effects of controlled substance use on personal health, safety, and the work environment, and about recognizing behavioral manifestations of controlled substance use or abuse. 49 C.F.R. § 391.121 .

C. Procedural History

Following promulgation of the regulations, the Unions filed a number of suits in the United States District Court for the Northern District of California challenging the validity of the regulations. In one of the cases, the district court granted a preliminary injunction staying the implementation of random and post-accident drug testing pending a determination of their constitutionality and the FHWA’s compliance with the dictates of the Administrative Procedure Act. Owner-Operators Indep. Drivers Ass’n v. Burnley, 705 F.Supp. 481 (N.D.Cal.1989). In issuing the injunction, the district court held that “the regulations providing for random drug testing and indiscriminate post-accident drug testing lack the particularized, reliable findings about *1293 drug use or impairment in the industry necessary to justify the intrusiveness of the search.” Id. at 485 (footnote omitted).

On November 6, 1989, the FHWA announced that, in light of the injunction issued in Owner-Operators, implementation of random and post-accident drug testing would be deferred until further notice. 54 Fed.Reg. 46,616. The FHWA noted that, if the injunction were dissolved on appeal, new implementation dates for random and post-accident testing would be promulgated. Id. at 46,617. As to the remaining types of testing, the original implementation dates remained in effect. Id.

Five suits challenging the validity of the regulations are before us today. In two of these cases, the International Brotherhood of Teamsters and the Amalgamated Transit Union agreed that the district court lacked jurisdiction to hear their petitions for review. A stipulated order transferring the case to this court was entered, pursuant to 28 U.S.C. § 1631 (1988). However, the district court included in the order the statement: “this court not having determined the issue of jurisdiction or expressing its opinion by approval of this Order.” Amalgamated Transit Union v. Skinner, No. C-89-0081-MHP (N.D.Cal. April 6, 1989) (order transferring cases). The two suits filed by the Railway Labor Executives’ Association likewise were transferred to this court, at the Association’s request, under 28 U.S.C. § 1631 . In the final case, the Oil, Chemical, and Atomic Workers Union (“OCAW”) opposed a transfer to this court. There the district court ruled that it lacked jurisdiction and transferred the case here, again pursuant to 28 U.S.C. § 1631 . These five cases were subsequently consolidated.

The district court in Owner-Operators reached a contrary result, concluding that it did have jurisdiction to hear the case and thus declining to transfer it to this court. The court did, however, certify its jurisdictional ruling for interlocutory appeal. On October 17, 1989, we granted the petition for interlocutory appeal.

JURISDICTION

We hold today, in Owner-Operators Indep. Drivers Ass’n v. Skinner, 931 F.2d 582 (9th Cir.1991), that courts of appeals have exclusive jurisdiction to hear petitions for review of final agency orders and regulations issued by the FHWA.

In its amicus brief filed in this case, Owner-Operators contends that we lack jurisdiction in four of these cases because the district courts did not expressly predicate their section 1631 case transfers on a finding of “want of jurisdiction.” 3 Owner-Operators urges us to remand these cases to the district courts without reaching the merits of the constitutional and Administrative Procedure Act claims. We decline to do so.

Under 28 U.S.C. § 1631 ,

Congress enacted section 1631 for the express purpose of eliminating the “unnecessary risk that a litigant may find himself without a remedy because of a lawyer’s error or a technicality of procedure.” S.Rep. No. 275, 97th Cong., 2d Sess. 11, reprinted in 1982 U.S.Code Cong. & Admin.News 11, 21.

*1294 Remanding four of these five cases would be both futile and contrary to the purpose of section 1631. First, we have uncontested authority to rule upon the petition for review filed in the case brought by OCAW. In that case, the district court explicitly found that it lacked jurisdiction prior to transferring the case under section 1631. The OCAW petition encompasses all of the constitutional and Administrative Procedure Act issues presented in the other four petitions. In deciding OCAW’s case, we necessarily dispose of the other four petitions as well. Remanding so that the district court may enter a predetermined ruling on jurisdiction and then transfer the case back to us so that we might enter a preordained ruling on the merits would be an exercise in futility and a waste of judicial resources.

Second, this court has previously declined to saddle section 1631 with the straitened interpretation proffered by Owner-Operators. See, e.g., Harris v. McCauley (In Re McCauley), 814 F.2d 1350, 1352 (9th Cir.1987) (“It would be a curious procedure to remand this case to the district court so that the district judge could decide whether or not to officially transfer it back to us.”). Congress intended section 1631 to promote the “interests] of justice.” Jurisdictional substance, rather than procedural niceties or magic words, governs the propriety of transfers under section 1631. In the four cases at issue, the district courts all refused to address the merits of the petitions and voluntarily transferred the cases to this court. The district courts did not, in fact, have jurisdiction over those cases. This court alone is empowered to hear them. We thus accepted the transfers. In the end, the petitions are in their proper forum and none of the jurisdictional boundaries prescribed by Congress has been transgressed. No purpose would be served by remand. Rather than engage in a game of “jurisdictional badminton,” 4 we hold that, in the interest of justice, we have jurisdiction to hear all five petitions. Accord Newman-Green, Inc. v. Alfonzo-Larrain, 490 U.S. 826, 837 , 109 S.Ct. 2218, 2225 , 104 L.Ed.2d 893 (1989) (a litigant “should not be compelled to jump through ... judicial hoops merely for the sake of hypertechnical jurisdictional purity”).

STANDARD OF REVIEW

We review the constitutionality of agency drug testing regulations de novo. See Bluestein v. Skinner, 908 F.2d 451, 454 (9th Cir.1990), cert. denied, — U.S. -, 111 S.Ct. 954 , 112 L.Ed.2d 1042 (1991). Because the Unions have challenged the regulations on their face, rather than as applied, we decide only the narrow question of whether these drug tests “can ever be conducted” without offending the fourth amendment. Skinner v. Railway Labor Executives’ Ass’n, 489 U.S. 602 , 632-33 n. 10, 109 S.Ct. 1402 , 1421 n. 10, 103 L.Ed.2d 639 (1989). “The challenge[ ] must establish that no set of circumstances exists under which the [regulations] would be valid.” United States v. Salerno, 481 U.S. 739, 745 , 107 S.Ct. 2095, 2100 , 95 L.Ed.2d 697 (1987).

We will find the agency’s decision to enact these regulations to be arbitrary and capricious only if the agency failed to consider relevant factors and has made a clear error in judgment. Sierra Pac. Indus. v. Lyng, 866 F.2d 1099 , 1105 (9th Cir.1989). The FHWA must carefully explain its reasons for the regulations and the regulations themselves must be reasonable. International Bhd. of Elec. Workers, Local 1245 v. Skinner, 913 F.2d 1454, 1457 (9th Cir.1990). A reasonable, albeit controversial, decision may not be overturned as arbitrary and capricious. Bluestein, 908 F.2d at 457 .

DISCUSSION

I. The Constitutionality of the Regulations

Precedent dictates much of the framework for our analysis. It is now *1295 established beyond peradventure that drug tests constitute searches within the meaning of the fourth amendment. Railway Labor, 489 U.S. at 617 , 109 S.Ct. at 1413 ; National Treasury Employees Union v. Von Raab, 489 U.S. 656, 665 , 109 S.Ct. 1384, 1390 , 103 L.Ed.2d 685 (1989); Bluestein, 908 F.2d at 455 . Testing conducted by private employers under compulsion of governmental regulations amounts to governmental action subject to the fourth amendment’s restrictions. Railway Labor, 489 U.S. at 614-16 , 109 S.Ct. at 1411-12 ; Bluestein, 908 F.2d at 455 .

A. The Applicability of the Warrant Requirement

We have recognized that the fourth amendment’s warrant requirement will not necessarily apply in the drug testing context. Bluestein, 908 F.2d at 455 ; see also Treasury Employees, 489 U.S. at 665-67 , 109 S.Ct. at 1390 .

Bluestein, 908 F.2d at 455 (quoting Treasury Employees, 489 U.S. at 665-66 , 109 S.Ct. at 1390 ).

In the present case, the drug-testing program does promote “special governmental needs, beyond the normal need for law enforcement.” Here, as in Railway Labor, Treasury Employees, and Bluestein, the agency implemented drug testing regulations to promote public safety and to deter drug use (with all of its concomitant costs) in the workplace. The FHWA emphasized, in enacting these regulations, that its aim was “to detect and deter the use of drugs by bus and truck drivers” and “to ensure a drug-free transportation environment which, in turn, will reduce accidents and casualties in motor carrier operations.” 53 Fed.Reg. at 47,135.

The regulations, moreover, preclude the tests’ use for law enforcement purposes by protecting the confidentiality of test results. Only the employer and the MRO learn of positive readings. The regulations strictly proscribe the disclosure of test results to third persons, such as law enforcement personnel, absent the consent of the driver. 49 C.F.R. § 391.89 . See Treasury Employees, 489 U.S. at 666 , 109 S.Ct. at 1391 ; Bluestein, 908 F.2d at 455 .

We must therefore determine whether a warrant is constitutionally required by balancing the drivers’ privacy expectations against the government’s interests. Previous drug-testing decisions advise that this balance weighs heavily against a warrant requirement. First, motor carrier officials, like the supervisors in Railway Labor, have little if any practical experience in the intricate processes of criminal investigation or warrant applications. Railway Labor, 489 U.S. at 623-24 , 109 S.Ct. at 1416-17 . “ ‘Imposing unwieldy warrant procedures ... upon supervisors, who would otherwise have no reason to be familiar with such procedures, is simply unreasonable.’ ” Id. (quoting O’Connor v. Ortega, 480 U.S. 709, 722 , 107 S.Ct. 1492, 1500 , 94 L.Ed.2d 714 (1987) (plurality)).

Second, the standardized nature of these tests and the minimal discretion afforded the administrators further undercut the need for a warrant in these circumstances. The regulations rigorously define how the tests are to be administered and carefully delimit the extent of the intrusion on individual privacy. The regulations also stipulate how drivers are to be selected for testing. With respect to preemployment, post-accident, and biennial drug testing, the drivers are selected by objectively discernible triggering events (a job application, accident, and passage of a preestablished period of time, respectively). The test administrators thus have no discretion over whom to test and when. Requiring a warrant for these tests thus would offer employees little additional protection because, given the “standardized nature of the tests and the minimal discretion vested in those charged with administering the program, there are virtually no facts for a neutral *1296 magistrate to evaluate.” Railway Labor, 489 U.S. at 622 , 109 S.Ct. at 1415-16 .

With respect to random testing, administrators concededly have somewhat more control over the timing of the tests. But they have no authority to determine which employees will be tested. According to the regulations, the selection must be made in a scientifically acceptable random manner. Administrators thus have no discretion to target individual employees for excessive testing.

The Unions raise the specter of excessively frequent or ill-timed administration of the drug tests. We are, however, confined to reviewing the facial validity of these regulations. We thus inquire only whether random testing could be administered constitutionally. Railway Labor, 489 U.S. at 632 -33 n. 10, 109 S.Ct. at 1421 n. 10. On their face, the regulations permit random testing at a rate of approximately one test per driver every two years. We cannot say that the limited discretion permitted employers to decide the timing of the tests means that they could never be administered constitutionally. Indeed, we have already held that random drug tests may proceed without a warrant. Bluestein, 908 F.2d at 457 . There will be time enough, if and when the case presents itself, to determine the constitutionality of these regulations in light of a particular employer’s abuses.

In sum, we hold that the fourth amendment does not mandate that motor carriers obtain a warrant prior to the administration of the drug tests ordered by the FHWA.

B. Random Drug Testing

Determining that no warrant is necessary to conduct these drug tests is only the first step in our inquiry into the constitutionality of these regulations, however. The drug tests at issue are conducted without either probable cause or individualized suspicion. While a search must ordinarily be based on probable cause, in certain circumstances the government’s interest “is sufficiently compelling to justify the intrusion on privacy entailed by conducting such searches without any measure of individualized suspicion.” Treasury Employees, 489 U.S. at 668 , 109 S.Ct. at 1392 . We find that such circumstances exist in this case.

1. Intrusion on privacy

The act of urination required to collect the drivers’ drug-testing specimen is an activity “traditionally shielded by great privacy.” Railway Labor, 489 U.S. at 626 , 109 S.Ct. at 1418 . But the inquiry we undertake here asks not whether a particular search constitutes, in the abstract or in general, a substantial intrusion on personal privacy. Rather, we must assess the incremental decrease in privacy caused by the search relative to the existing privacy expectations of the drivers. See Railway Labor, 489 U.S. at 624-25 , 109 S.Ct. at 1416-17 (privacy expectations defined in terms of “additional interference” on top of existing restrictions on employees); Treasury Employees, 489 U.S. at 671-72 , 109 S.Ct. at 1393 .

The privacy expectations of commercial truck drivers are markedly less than those of the public in general. The trucking industry is highly regulated and drivers have long been subjected to federal regulation of their qualifications. See, e.g., 49 U.S.C. § 3102 (b) (authorizing Secretary of Transportation to prescribe qualifications of drivers in order to promote safety in the motor carrier industry); 49 C.F.R. § 391.41 (regulating physical condition of drivers).

The regulation of drivers’ qualifications already includes subjecting them to comprehensive biennial physical examinations. 49 C.F.R. 391.43. The scope of the examination prescribed by the regulations. These examinations include urinalysis to detect infections of the genito-urinary tract, a history of luetic infection, or latent syphilis. 5 Drivers also must take road *1298 tests and written examinations. 49 C.F.R. §§ 391.31-391-37 . Drivers’ hours of service are strictly regulated in order to prevent fatigue. 49 C.F.R. § 395.3 .

Accordingly, the intrusiveness of these drug-testing regulations, on their face, must be measured against the impositions on drivers’ privacy already worked by the nature of their job and its attendant regulations. Indeed, drivers must maintain logs documenting when they are off-duty or in their sleeper berths. 49 C.F.R. § 395.8 . In Treasury Employees, the Supreme Court specifically noted that employees’ privacy expectations necessarily would be lowered if their positions already entailed “background investigations, medical examinations, or other [such] intrusions.” Treasury Employees, 489 U.S. at 677 , 109 S.Ct. at 1396 ; see also National Fed’n of Fed. Employees v. Cheney, 884 F.2d 603 , 612-13 (D.C.Cir.1989), cert. denied, — U.S. -, 110 S.Ct. 864 , 107 L.Ed.2d 948 (1990) (privacy expectations of security guards reduced because they are already subjected to extensive medical examinations and background checks); Thomson v. Marsh, 884 F.2d 113, 115 (4th Cir.1989) (employees at chemical plant enjoy reduced privacy expectations vis-a-vis drug testing because they are already subject to annual physicals that include urine testing); Schaill by Kross v. Tippecanoe County School Corp., 864 F.2d 1309, 1318 (7th Cir.1988) (drug testing of student athletes upheld, in part, because privacy expectations were reduced due to the fact that athletes already had to provide urine samples in the course of their routine physical examinations).

On the face of the regulations, motor carriers may limit their random drug tests so that they are, on average, no more frequent than the biennial exams. See 49 C.F.R. § 391.93 (d) (random testing may be at a rate of fifty percent per year). The design of the regulations, moreover, limits the intrusion on privacy as much as practicable by permitting unobserved urination, requiring that monitors be trained in specimen collection, and treating test results confidentially. We note that the testing procedures adopted by the FHWA closely track the procedures sanctioned by the Supreme Court in Treasury Employees, 489 U.S. at 672 -73 n. 2, 109 S.Ct. at 1394 n. 2. In light of the fact that (i) commercial drivers’ operate in a highly regulated industry, (ii) have submitted to substantial federal monitoring of their physical health and qualifications, and (iii) already must undergo urinalysis biennially, we conclude that the incremental decrease in privacy occasioned by the collection of a second urine specimen for drug-testing is, in this narrow context, constitutionally tolerable. Cf. Bluestein, 908 F.2d at 455-57 . That the drug testing procedure erected by the regulations is solicitous, to the extent possible, of drivers’ privacy concerns further buttresses our holding. See Railway Labor, 489 U.S. at 626-27 , 109 S.Ct. at 1418 *1299 (procedures for specimen collection minimize intrusion on privacy).

The Unions contend that the timing of random drug tests under the FHWA regulations makes the intrusion on privacy constitutionally prohibitive. The Unions’ objection targets two aspects of the random testing program: the surprise inherent in random testing and the repetition of the tests over the course of a driver’s career.

With respect to the element of surprise, we have recognized that the lack of notice inherent in a random testing program “is a relevant consideration; and, in a particularly close case, it is possible that this factor would tip the scales.” Bluestein, 908 F.2d at 457 (quotation omitted). We conclude, however, that the element of surprise does not “tip the scales” in this instance.

First, we note that the measure of surprise under the fourth amendment is not the apprehension of one who has been using drugs, but rather the anxiety “engendered in law abiding [drivers] by the nature of the” search. Michigan Dep’t of State Police v. Sitz, — U.S. -, 110 S.Ct. 2481, 2486 , 110 L.Ed.2d 412 (1990). While random testing by definition involves some element of surprise, the amount of anxiety should not be substantial. Drivers will be aware of the existence of a random drug-testing scheme, so while the precise time of the test will be unknown, the fact that they are subject to this search procedure will not be a surprise. Cf. Schaill by Kross, 864 F.2d at 1322 (use of published selection criteria reduced stigma and fear associated with selection for drug testing). The uncertainty confronting the drivers is thus unlike the surprise occasioned by roving border patrols, which the Supreme Court found constitutionally intolerable in Almeida-Sanchez v. United States, 413 U.S. 266 , 93 S.Ct. 2535 , 37 L.Ed.2d 596 (1973), where citizens did not know in advance that a search would be forthcoming. Rather, this situation is more comparable to the traffic checkpoints upheld in United States v. Martinez-Fuerte, 428 U.S. 543, 558-59 , 96 S.Ct. 3074, 3083 , 49 L.Ed.2d 1116 (1976). Here, as there, the persons being searched will know that others are subject to the same search procedure and will have visible indicia of the search administrator’s authority. Cf. id.; see also United States v. Ortiz, 422 U.S. 891, 894-95 , 95 S.Ct. 2585, 2587-88 , 45 L.Ed.2d 623 (1975).

Second, this court has previously upheld random searches, despite the element of surprise they contain. Bluestein, 908 F.2d at 456-57 . Nothing suggests that the element of surprise works commercial drivers any greater hardship than it did the airline personnel in Bluestein.

Moreover, in sustaining post-accident drug testing in Railway Labor, the Supreme Court implicitly acknowledged that searches may be conducted without advance notice and yet remain faithful to the fourth amendment, at least as long as the surprise is necessary to accomplish the government’s purpose. Railway Labor, 489 U.S. at 630 , 109 S.Ct. at 1420 (“By ensuring that employees in safety-sensitive positions know they will be tested upon the occurrence of a triggering event, the timing of which no employee can predict with certainty, the regulations significantly increase the deterrent effect....”) (emphasis added); see also New York v. Burger, 482 U.S. 691, 710 , 107 S.Ct. 2636, 2648 , 96 L.Ed.2d 601 (1987) (unannounced administrative searches of automobile junkyards upheld despite element of surprise, because surprise was “crucial” to attainment of regulatory goals). Train accidents are no more predictable than the date of a random drug test.

In sum, we hold that, on the circumstances of this case, the privacy intrusion effected by random searches is minimal, given the commercial drivers’ already substantially reduced privacy expectations.

2. Compelling governmental interest

We must balance against the intrusion on drivers’ privacy the strength of the governmental interest inspiring the institution of random drug testing. Bluestein, 908 F.2d at 455 . The FHWA cited enhanced transportation safety, accident avoidance, and deterrence of drug use as its primary motivation for enacting these *1300 regulations. 53 Fed.Reg. at 22,271. The Unions argue that these reasons are not sufficiently compelling to justify an abridgement of the drivers’ fourth amendment rights. They stress the lack of evidence of a serious drug problem among commercial drivers and the FHWA’s failure to rely on less intrusive alternatives to drug-testing to achieve its goals. We reject both arguments and hold that the FHWA has enunciated a compelling governmental interest in conducting random drug tests of commercial drivers.

The FHWA has a compelling interest in preventing drivers from using illegal drugs while behind the wheel. Both this court and the Supreme Court have acknowledged the vital governmental interest in ensuring the sobriety and fitness of operators of dangerous instrumentalities or equipment. In Railway Labor, the Supreme Court found compelling the Department of Transportation’s interest in testing railway employees, including train operators, because such employees “discharge duties fraught with such risks of injury to others that even a momentary lapse of attention can have disastrous consequences” and make judgments that “can cause great human loss before any signs of impairment become noticeable to supervisors or others.” Railway Labor, 489 U.S. at 628 , 109 S.Ct. at 1419 .

Similarly, in Treasury Employees, the Supreme Court upheld drug testing for customs service employees who carry firearms. In so holding, the Court focused on the “extraordinary safety and national security hazards that would attend the promotion of drug users to positions that require the carrying of firearms.” Treasury Employees, 489 U.S. at 674 , 109 S.Ct. at 1395 .

In a case closely analogous to the one at hand, we recognized the Department of Transportation’s compelling interest in randomly testing airline employees, including pilots, for drug use. Bluestein, 908 F.2d at 456 . There, as here, the substantial harm that an airplane crash could occasion and the need to protect the safety of the traveling public prompted the promulgation of the random drug testing regulation. Id.; see also Electrical Workers, 913 F.2d at 1462 (“[T]he consequent harm that would occur from a pipeline accident is sufficient to merit the finding of a strong governmental interest in the detection and deterrence of substance abuse among pipeline workers.”).

The concern for safety and deterrence exhibited by the FHWA at least parallels, if not exceeds, the governmental interests found compelling in these earlier cases. A 26,000 pound truck, like the locomotive in Railway Labor and the airplane in Blue-stein, “becomes lethal when operated negligently by persons who are under the influence of ... drugs.” Railway Labor, 489 U.S. at 628 , 109 S.Ct. at 1419 (quoting Railway Labor Executives’ Ass’n v. Burnley, 839 F.2d 575, 593 (9th Cir.1988) (Alarcon, J., dissenting)). Given the enormous size of commercial trucks, relative to other vehicles on the road, a single mistake in judgment or momentary lapse in attention can have devastating consequences for other travelers. This is especially true for drivers who carry hazardous cargo. While a single accident may not imperil as many lives as a single airline or train crash, the vast number of drivers on the road at any given time 6 multiplies the danger to motorists and raises the FHWA’s concern for transportation safety to the level of a compelling governmental interest.

The Unions argue that the lack of hard evidence revealing a serious drug abuse problem among professional drivers belies the FHWA’s professed safety concerns. They dismiss as “anecdotal” the evidence garnered by the FHWA in support of the regulations.

The Unions’ objection does not diminish the strength of the governmental interest for two reasons. First, the administrative record reveals a reasonable basis for the FHWA’s concerns. Witnesses testified *1301 about their personal knowledge of drug use by commercial drivers. Others noted that the nature of the profession (long hours, long periods alone and away from home) heightens drivers’ susceptibility to drug use and abuse. Statistical studies documenting drug usage by drivers and revealing the deterrent effect of drug testing were also introduced. While reasonable people may debate the weight to be accorded various types of evidence, the FHWA nonetheless had a reasonable basis for deciding to implement drug testing.

Furthermore, even assuming no evidence of substantial drug abuse among truck drivers existed, the FHWA’s interest in deterring drug abuse and in “preventing an otherwise pervasive societal problem from spreading” to the trucking industry would “furnish[] an ample justification” for the testing program. Treasury Employees, 489 U.S. at 674 -75 & n. 3, 109 S.Ct. at 1395 & n. 3. Indeed, we noted in Bluestein that nothing in Supreme Court precedent predicates the constitutionality of a drug testing program on the preexistence of a drug abuse problem in the group to be tested. Bluestein, 908 F.2d at 456 n. 6; see also Electrical Workers, 913 F.2d at 1461 (random testing of pipeline workers upheld despite the agency’s inability to document a widespread drug problem in the pipeline industry). “[T]he deterrent purposes of the [FHWA’s] program and the potential for serious harm” constitute a compelling justification for the regulations. Bluestein, 908 F.2d at 456 .

The Unions also contend that the governmental interest is not compelling because alternative, less intrusive means exist to promote transportation safety and drug deterrence. The Supreme Court, however, has explicitly rejected the notion that the constitutionality of a drug testing program turns upon the availability of less intrusive means to achieve the agency’s goals. Railway Labor, 489 U.S. at 629 n. 9, 109 S.Ct. at 1419 n. 9 (reasonableness of post-accident drug testing “ ‘does not necessarily or invariably turn on the existence of alternative “less intrusive” means’ ”) (quoting Illinois v. Lafayette, 462 U.S. 640, 647 , 103 S.Ct. 2605, 2610 , 77 L.Ed.2d 65 (1983)).

Furthermore, it is not clear that alternative types of drug tests, such as reasonable suspicion testing, would accomplish the dual goals of safety and deterrence. The FHWA has enacted a comprehensive drug testing program involving a variety of different drug tests. In its considered judgment, such a thoroughgoing approach was necessary to ensure that commercial drivers abstain from drug use throughout their careers. Random drug testing has particular appeal because, due to its inherent element of surprise, complete abstinence from drug use is the drivers’ only safeguard against failure. Temporary abstinence will often suffice to survive periodic or preem-ployment drug screens. Post-accident testing, although also containing an element of surprise, does not pack as powerful a deterrent punch because it is less frequent and less certain.

Random testing also responds to the “operational realities of the [drivers’] workplace.” O’Connor v. Ortega, 480 U.S. at 717 , 107 S.Ct. at 1497 . Commercial drivers work alone and unsupervised for substantial periods of time. The opportunity for supervisors to detect drug usage, and thus have reasonable suspicion to conduct a test, is minimal. As a result, drivers could cause “great human loss before any signs of impairment become noticeable to supervisors or others.” Railway Labor, 489 U.S. at 628 , 109 S.Ct. at 1419 .

Given the unique deterrent effect of random testing and the ineffieaey of relying on individualized suspicion in this particular employment context, the FHWA reasonably found the alternative, less intrusive drug tests to be insufficient to meet its goals. We are ill-equipped to second-guess the FHWA’s judgment on this matter. Railway Labor, 489 U.S. at 629 n. 9, 109 S.Ct. at 1419 n. 9.

3. The balance

Having concluded that the incremental intrusion on drivers’ privacy worked by random drug testing is minimal and that the FHWA’s purpose for instituting the *1302 tests is compelling, we conclude that random drug testing of commercial drivers does not offend the fourth amendment despite the lack of a warrant or individualized suspicion. Reinforcing our decision to strike this balance are the recent opinions of numerous courts, including our own, upholding random drug testing under similar circumstances. See, e.g., Electrical Workers, 913 F.2d at 1464 (random drug testing of pipeline workers); Hartness v. Bush, 919 F.2d 170, 173 (D.C.Cir.1990) (random testing of government employees with “secret” national security clearances); Bluestein, 908 F.2d at 455-57 (random testing of aviation personnel); Taylor v. O’Grady, 888 F.2d 1189 , 1199 (7th Cir.1989) (random testing of correctional officers who have contact with prisoners); American Fed’n of Gov’t Employees v. Skinner, 885 F.2d 884, 889-93 (D.C.Cir.1989), cert. denied, — U.S. -, 110 S.Ct. 1960 , 109 L.Ed.2d 321 (1990) (random testing of transportation employees); National Fed’n of Fed. Employees, 884 F.2d at 610-13 (random testing of certain Army civilian employees); Thomson v. Marsh, 884 F.2d at 115 (random testing of employees at chemical weapons plant); Harmon v. Thornburgh, 878 F.2d 484, 496 (D.C.Cir.1989), cert. denied, — U.S. -, 110 S.Ct. 865 , 107 L.Ed.2d 949 (1990) (random testing of Justice Department employees with “top secret” clearance); Guiney v. Roache, 873 F.2d 1557, 1558 (1st Cir.), cert. denied, — U.S. -, 110 S.Ct. 404 , 107 L.Ed.2d 370 (1989) (random testing of police officers); Policeman’s Benevolent Ass’n, Local 318 v. Township of Washington, 850 F.2d 133, 141 (3d Cir.1988), ce rt. denied, 490 U.S. 1004 , 109 S.Ct. 1637 , 104 L.Ed.2d 153 (1989) (random testing of police officers); Rushton v. Nebraska Public Power Dist., 844 F.2d 562, 567 (8th Cir.1988) (random testing of nuclear power plant employees); McDonnell v. Hunter, 809 F.2d 1302 , 1308-09 (8th Cir.1987) (random testing of correctional officers). We would be hard-pressed to find the transportation safety concerns articulated by the FHWA any less compelling or the privacy intrusion any more substantial in this instance than in any of the foregoing cases. We therefore hold that the FHWA’s regulations mandating random drug testing are constitutional on their face.

C. Periodic Testing

The Unions also raise fourth amendment objections to the institution of periodic (biennial) drug testing. Under the regulations, motor carriers must conduct a drug test on all drivers during their first medical examination following commencement of the entire drug testing program. 49 C.F.R. § 391.105 (a). 7 A year after random drug testing has been fully implemented, motor carriers may discontinue periodic urinalysis for controlled substance use. 49 C.F.R. § 391.105 (c).

The Unions argue here, as they did with random testing, that the intrusion on drivers’ privacy outweighs the governmental purposes accomplished by periodic testing. They also argue that the regulations’ failure to require motor carriers to discontinue periodic testing after random testing commences exacerbates the constitutional violation because random testing will have already accomplished the goals of safety and deterrence. Periodic testing will be incapable of further advancing these ends. We find these arguments unpersuasive and affirm the validity of the periodic testing regulations.

The constitutional analysis we applied to upholding random drug testing applies with equal vigor here. In fact, the drivers’ privacy objections have significantly less force in this context. The testing will take place once, during the course of a thorough physical examination that already entails urinalysis. Moreover, periodic testing does not involve the element of surprise that the Unions found so troublesome with random drug testing. The incremental increase in intrusiveness of asking drivers to contribute two, rather than one, specimen in the course of their medical examinations does not tip the constitutional scales against permitting the search.

On the other hand, the government’s interest in conducting the tests is not de *1303 creased. Because random testing could, by definition, miss some drivers, the FHWA has a compelling interest in guaranteeing that all drivers are tested at least once. Blanket testing will better promote transportation safety and will further advance deterrence, because no driver using or contemplating using drugs will be tempted to play the odds in the hope that random testing does not target him. All will know that all will be tested.

While the constitutional balance might be struck differently if the FHWA mandated continued periodic testing even after random drug testing is in full swing, this is not such a case. The FHWA specifically permits the early termination of biennial testing to avoid redundancy. To the extent private motor carriers may elect, at that point, to continue periodic testing, they will no longer be doing so at the behest of the federal government and the fourth amendment will thus no longer constrain their conduct. Railway Labor, 489 U.S. at 614 , 109 S.Ct. at 1411 (“[T]he Fourth Amendment does not apply to a search or seizure, even an arbitrary one, effected by a private party on his own initiative.”).

D. Pre-Employment Testing

In addition to random and biennial drug screens, the regulations mandate that motor carriers administer drug tests to all prospective employees whom they intend to hire. 49 C.F.R. § 391.103 (a). 8 The Unions articulate essentially the same objections to pre-employment testing that they raise with respect to random and periodic testing. For the same reasons we sustained the constitutionality of the latter two tests, we hold that pre-employment testing does not violate the fourth amendment.

The privacy expectations of people seeking to work as commercial drivers are significantly reduced relative to other members of society. They have voluntarily chosen to enter a highly regulated profession that already requires periodic extensive physical examinations and urinalysis to determine the qualifications of its members. In Treasury Employees, the Supreme Court held that persons seeking new employment positions in a highly regulated profession involving compelling public safety and security concerns could be subjected to drug tests prior to starting their jobs. Treasury Employees, 489 U.S. at 677 , 109 S.Ct. at 1396 (Customs employees tested upon promotion to sensitive positions). In reaching this conclusion, the Court attached great significance to the fact that the new positions already required background investigations and medical examinations. Id. (“We also agree that employees who seek promotions to positions where they would handle sensitive information can be required to submit to a urine test under the Service’s screening program, especially if the positions covered under this category require background investigations, medical examinations, or other intrusions that may be expected to diminish their expectations of privacy in respect of a urinalysis test.”) (emphasis added).

The level of intrusion, furthermore, is less than that encountered in random testing. No element of surprise is involved. The test is triggered by the job applicant’s own voluntary conduct, and will occur only once in the applicant’s career.

The FHWA’s compelling interest in transportation safety justifies this narrow privacy intrusion. Unlike random or periodic testing, pre-employment testing ensures that drivers who use drugs never get behind the wheel of a 26,000 pound truck in the first instance. The FHWA need not wait until a driver is actually working (and thus a potential daily hazard) to begin promoting safety and deterrence. The regulations prevent the transportation risk from ever coming to fruition. 9 This concern for preempting accidents outweighs the minimal intrusion on job applicants’ privacy, making the conduct of suspicionless, pre-employment testing constitutional.

*1304 E. Post-Accident Testing

The fourth and final constitutional objection levelled against the FHWA’s regulations concerns the requirement that drivers arrange on their own to be tested for drug use within thirty-two hours of a “reportable accident.” 49 C.F.R. § 391.113 (a). The regulations consider an accident “reportable” if it involves (1) a fatality, (2) an injury demanding immediate medical treatment away from the scene of the accident, or (3) at least $4,400 in property damage. 49 C.F.R. § 394.3 . 10

The Unions’ objections to post-accident drug testing echo the concerns they voiced with respect to random, periodic, and post-accident testing (such as the extent of the intrusion on privacy and the lack of evidence that drug use is a significant factor in accidents involving commercial trucks). We reject these arguments and hold the regulations constitutional.

The Supreme Court has already sustained the constitutionality of post-accident testing for railroad operators. Railway Labor, 489 U.S. at 626-32 , 109 S.Ct. at 1418-21 . In so holding, the Court stressed the deterrent value of such tests.

Id. at 630 , 109 S.Ct. at 1420 (citation omitted). The Court also noted that such tests would help the railroads “obtain invaluable information about the causes of major accidents ... and to take appropriate measures to safeguard the general public.” Id. (citation omitted).

Those same policy concerns obtain in this case. The FHWA cited both deterrence and the need to gather information about the causes of accidents as its basis for promulgating the post-accident drug testing regulations. 53 Fed.Reg. at 47,140-41. At the same time, nothing in the record suggests that the commercial drivers’ expectation of privacy is any greater than that enjoyed by the railroad employees in Railway Labor. Indeed, the drivers’ privacy expectations may be less because nothing in Railway Labor indicates that railway employees were already required to undergo biennial urinalysis. We accordingly uphold the constitutionality of the post-accident drug testing regulation. 11

II. The Arbitrary and Capricious Claim

In addition to their constitutional objections, the Unions assert that the FHWA acted arbitrarily and capriciously in enacting these regulations. They object, in particular, to the lack of an evidentiary basis for believing either that a drug problem exists in the industry or that the conduct of drug tests will cure the problem.

We hold that the promulgation of these regulations was not arbitrary and capricious. The agency specifically explained why it chose to implement these regulations. It collected evidence, both testimonial and empirical, documenting a drug use problem in the motor carrier industry and demonstrating the efficacy of these various drug tests in achieving the goals of deterrence and transportation safety. The agency also considered how these tests would affect drivers’ privacy, as evidenced by its adoption of testing procedures designed to maximize driver privacy and confidentiality. While the FHWA’s investigation might not have been paradigmatic, it nonetheless *1305 was sufficiently reasonable to justify its actions. See Bluestein, 908 F.2d at 457 .

CONCLUSION

Today we uphold a massive drug testing program that will touch the lives of literally millions of citizens. We do not do so lightly. We share many of the Unions’ concerns about the substantial inroads drug testing makes on our precious fourth amendment freedoms. But we do not write upon a clean slate in this area. Much of our decision is compelled by prior decisions of the Supreme Court and this Circuit. Unless and until Congress or the Supreme Court reconsiders the enormous constitutional cost, in terms of lost privacy, dignity, and autonomy, resulting from the “war on drugs,” we are bound to apply the law as it exists. We therefore hold that, given the comprehensive governmental regulation to which commercial drivers are already subject, the FHWA’s random, biennial, pre-employment, and post-accident drug testing regulations are constitutional on their face.

PETITION FOR REVIEW DENIED.

. During the first year of implementation, motor carriers need test only at a rate equal to 25 percent of the eligible drivers. 49 C.F.R. § 391.93 (d)(3).

. The regulation lists four indicia warranting monitored urination: (1) abnormal temperature of the specimen, (2) the previous specimen had an abnormally low specific gravity or creatinine concentration, (3) clear and unequivocal evidence of an attempt to substitute or adulterate the sample, and (4) rehabilitation or follow-up testing of an employee previously identified as using a controlled substance. 49 C.F.R. § 40.25 (e)(2). Moreover, the concurrence of either a supervisor of the collection site monitor or an employer representative must be obtained prior to any direct observation of urination. 49 C.F.R. § 40.25 (e)(3).

. Owner-Operators does not contend that we lack jurisdiction over the suit filed by OCAW (No. 89-70248). It challenges only our authority to hear the actions filed by the International Brotherhood of Teamsters (No. 89-70165), the Amalgamated Transit Union (No. 89-70166), and the Railway Labor Executives’ Association (Nos. 89-70185, -70186).

. Investment Co. Inst. v. Board of Governors of the Fed. Reserve Sys., 551 F.2d 1270, 1283 (D.C.Cir.1977) (Leventhal, J., concurring) (quoting Natural Resources Defense Council, Inc. v. Environmental Protection Agency, 512 F.2d 1351, 1361 (D.C.Cir.1975) (Skelly Wright, J., dissenting in part)).

. The scope of the examination prescribed by the regulations evidences the substantial intrusion on their physical privacy already accepted by commercial drivers:

. The FHWA estimates that approximately three million drivers will be tested under its program. 53 Fed.Reg. at 47,149.

. As noted earlier, drivers are already required to undergo biennial medical examinations, including urinalysis, to determine if they are fit to drive. 49 C.F.R. §§ 391.41 , 391.43.

. The regulations except drivers who are regular employees of another carrier (and thus presumably have already been tested under that motor carrier’s program) or who demonstrate that they already participate in a comparable drug testing program. 49 C.F.R. § 391.103 (d).

. We note that the FHWA’s decision to require preemployment testing was informed, at least in *1304 part, by record evidence of high drug use rates among job applicants. A study conducted by Greyhound Bus Lines in 1983 revealed that around 30 percent of its prospective employees tested positive for controlled substances.

. A driver who is seriously injured in the accident and, as a result, is unable to provide a urine specimen must subsequently authorize the release of medical records that would indicate whether controlled substances were found in her or his system. 49 C.F.R. § 391.113 (b).

. We express no opinion on whether the government’s interest in data collection alone would outweigh the intrusion on drivers’ privacy in this instance. See Electrical Workers, 913 F.2d at 1461 n. 19.