Peter A. Lust brought a personal injury action against Merrell Dow Pharmaceuticals, Inc. alleging that his birth defect, diagnosed as hemifacial microsomia, was caused by his mother’s ingestion of Clomid, a fertility drug manufactured by Merrell Dow. California law dictates that causation is a necessary element of a personal injury action. The district court granted summary judgment in favor of Merrell Dow on the grounds that Lust’s only causation evidence — the expert opinion of Dr. Alan Done — did not reflect scientific knowledge and was inadmissible under F.R.E. 702. The district court had diversity jurisdiction, 28 U.S.C. § 1332 , and we have jurisdiction over the appeal. 28 U.S.C. § 1291 . We affirm.

*596 I. Background

The district court held an in limine hearing pursuant to F.R.E. 104(a) to determine whether Done’s opinion was admissible under F.R.E. 702. Done explained the reasoning underlying his conclusion that Clomid is a human teratogen (i.e., an agent that causes birth defects) that causes hemifacial microso-mia:

Done found that human epidemiological studies reported a positive association between ingestion of Clomid and a wide variety of birth defects, that animal studies reported Clomid to be teratogenic in four animal species, and that other studies reported Clomid to have a mutagenic affect on humans. He concluded that “since ... Clomid is the known specific drug capable of causing all kinds of birth defects, depending on the timing, it should be able to include [hemifacial microsomia].” He stressed that he had first expressed his theories on Clomid in a 1984 article that predates this litigation.

Merrell Dow elicited a number of concessions from Done on cross-examination. He admitted that he is not a teratologist, geneticist, or embryologist. He admitted that his 1984 article was not peer-reviewed and was based on research conducted in preparation for expert testimony in a different case concerning a different drug manufactured by Merrell Dow. He admitted that no peer-reviewed article concludes that Clomid is a human teratogen. And he admitted that no human epidemiological or animal studies have found a positive association between the ingestion of Clomid and hemifacial microso-mia.

Merrell Dow presented its own expert who testified that there were flaws in Done’s methodology:

Done responded by pointing to some articles that, he argued, supported his premise that a positive association between an agent and a wide variety of birth defects establishes that the agent substantially increases the probability of all types of birth defects.

The district court excluded Done’s testimony. It explained:

Since Lust had no causation evidence other than Done’s opinion, the district court granted Merrell Dow’s motion for summary judgment.

II. Causation

We review a grant of summary judgment de novo, Warren v. City of Carlsbad, 58 F.3d 439, 441 (9th Cir.1995), cent. denied, - U.S. -, 116 S.Ct. 1261 , 134 L.Ed.2d 209 (1996), and an F.R.E. 702 ruling for an abuse of discretion. See Daubert v. Merrell Dow Pharmaceuticals, Inc., 43 F.3d *597 1311, 1315 (9th Cir.), cert. denied, - U.S. 116 S.Ct. 189 , 133 L.Ed.2d 126 (1995). The abuse of discretion standard applies to an F.R.E. 702 ruling even though the ruling was dispositive of a motion for summary judgment. See id.; Rosen v. Ciba-Geigy Corp., 78 F.3d 316, 318 (7th Cir.1996).

F.R.E. 702 provides that “[i]f scientific ... knowledge will assist the trier of fact ... to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise.” In Daubert v. Merrell Dow Pharmaceuticals, Inc., the Supreme Court held that F.R.E. 702 did not enact Frye v. United States, 293 F. 1013, 1014 (D.C.Cir.1923), which first stated the rule that expert scientific opinion is inadmissible if it is based on a method that has not gained “general acceptance” in the relevant scientific community. Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 587-89 , 113 S.Ct. 2786, 2794 , 125 L.Ed.2d 469 (1993). Under the Daubert interpretation of F.R.E. 702, general acceptance is not a necessary condition to admissibility; expert scientific opinion is admissible if it qualifies as “scientific knowledge” and is therefore sufficiently “reliable.” Id. at 590 , 113 S.Ct. at 2794-95 .

Daubert maintained that district court judges would continue to act as gatekeepers, and it listed four nonexciusive factors that would assist in the separation of inadmissible opinions based on junk science from admissible opinions developed by the scientific method. District court judges are to consider not only (1) whether the method has gained general acceptance in the relevant scientific community, but also (2) whether the method has been peer-reviewed, (3) whether the method "can be (and has been) tested," and (4) whether there is a "known or potential rate of error." Id. at 594 , 113 S.Ct. at 2796-97 . The last two factors-testing and rate of error-do not apply, however, when the expert has not done original research, but rather has surveyed available literature and drawn conclusions that differ from those presented by the scientists who performed the original work. See Daubert, 43 F.3d at 1317 n. 4.

We have noted that the Daubert inquiry is flexible and have provided additional guidance. “One very significant fact” is whether the expert has “developed [his] opinions expressly for purposes of testifying,” since “a scientist’s normal workplace is the lab or the field, not the courtroom or the lawyer’s office.” Daubert, 43 F.3d at 1317 . That the expert failed to subject his method to peer-review and to develop his opinion outside the litigation is not dispositive, but if these guarantees of reliability are not satisfied, the expert “must explain precisely how [he] went about reaching [his] conclusions and point to some objective source ... to show that [he has] followed the scientific method, as it is practiced by (at least) a recognized minority of scientists in [his] field.” Id. at 1318-19 .

We begin our inquiry by determining whether Done’s research satisfies the applicable guarantees of reliability. Lust does not contend that Done subjected any of his research to peer review, but he does contend that Done developed his opinions on Clomid outside the litigation context. We reject the argument. Although Done published the 1984 article prior to this litigation, he was at that time already a professional plaintiffs witness. It is not unreasonable to presume that Done’s opinion on Clomid was influenced by a litigation-driven financial incentive.

Done could have convinced the district court that his methodology was nevertheless scientific if he had explained how he went about reaching his conclusions and had pointed to an objective source demonstrating that his method and premises were generally accepted by or espoused by a recognized minority of teratologists. But he failed to do so. Despite his claim that he relied on a generally accepted method, Done’s chief premise was that if there is evidence of a positive association between an agent and a wide variety of birth defects in human epidemiological and animal studies, then the agent substantially increases the probability of all types of birth defects. Merrell Dow’s expert testified to the effect that this premise is not espoused by a relevant minority of teratolo- *598 gists, and the articles Done submitted do not suggest otherwise.

We must address Done’s final contention that the district court violated Daubert’s command that “[t]he focus ... must be solely on principles and methodology, not on the conclusions that they generate.” 509 U.S. at 595 , 113 S.Ct. at 2797 . Done’s conclusions did arouse the district court’s suspicion, but that is to be expected. When a scientist claims to rely on a method practiced by most scientists, yet presents conclusions that are shared by no other scientist, the district court should be wary that the method has not been faithfully applied. It is the proponent of the expert who has the burden of proving admissibility. To enforce this burden, the district court can exclude the opinion if the expert fails to identify and defend the reasons that his conclusions are anomalous.

There was no abuse of discretion in the ruling that Lust failed to carry the burden of proving that Done’s testimony was admissible under F.R.E. 702. Summary judgment was appropriate since without Done’s testimony, Lust offered no evidence of causation, a necessary element of his personal injury action.

III. Celotex

Lust contends that the district court erred in granting summary judgment without determining whether expert affidavits filed with Merrell Dow’s motion for summary judgment would have been admissible. The Supreme Court held in Celotex Corp. v. Catrett that F.R.C.P. 56 contains no express or implied requirement that the moving party “support its motion with affidavits or other similar materials negating the opponent’s claim.” 477 U.S. 317, 323 , 106 S.Ct. 2548, 2553 , 91 L.Ed.2d 265 (1986). In Daubert , we rejected an argument identical to Lust’s:

Daubert, 43 F.3d at 1315 . The eases cited by Lust from other circuits are not in conflict. Merrell Dow satisfied its initial burden when it indicated that Lust’s only causation evidence might be inadmissible.

AFFIRMED.