This case marks the convergence of epidemiological evidence, probabilistic causation in carcinogenic torts, and the important issue of the extent to which a trial court may assess the sufficiency of scientific evidence, in light of the Supreme Court’s recent holding in Daubert v. Merrell Dow Pharmaceuticals, Inc., — U.S. -, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993). That decision enlarged district courts’ “gatekeeping” roles in appraising the admissibility of scientific evidence. The central question before us is the standard governing federal judges’ evaluations of the sufficiency — as opposed to admissibility — of scientific evidence already admitted.
In 1983, John Maiorana (“Maiorana”) died of colon cancer. His widow, plaintiff/appellant Arlene M. Maiorana (“plaintiff”), claimed that her husband’s illness was caused by exposure to Cafco D, an asbestos spray manufactured by defendant-appellee United States Mineral Products Co. (“USMP”). This spray was used for insulation on two construction sites — the World Trade Center in New York City and Meadowbrook Hospital in Nassau County, New York — where Maiorana was employed as a sheet metal worker.
The scientific community is divided on whether asbestos exposure significantly increases the risk of contracting colon cancer. At trial in the United States District Court for the Southern District of New York (Robert W. Sweet, Judge), both plaintiff and USMP brought expert witnesses and numerous epidemiological studies to bear on their dispute over the causal link between asbestos and colon cancer. Plaintiff also introduced into evidence Maiorana’s medical records and personal history in order to eliminate other likely causal factors. After considering this evidence, the jury on February 10, 1993, returned a verdict in favor of the plaintiff.
In an opinion dated July 23, 1993, the district court granted USMP’s motion for judgment as a matter of law, setting aside the jury verdict. See In re Joint E. & S. Dist. Asbestos Litig., 827 F.Supp. 1014 (S.D.N.Y.1993) (“Asbestos Litig. III”). We believe that the district court overstepped the boundaries of the role contemplated by Daubert and inappropriately usurped the role of the jury. Therefore, we reverse the order entering judgment as a matter of law for USMP, and we direct the reinstatement of the jury verdict in favor of plaintiff. We affirm those pre-trial rulings which third-party defendants/eross-appellants now challenge on appeal. Finally, we remand for further proceedings consistent with this opinion.
I. BACKGROUND
A. Factual History
Cafco D is a fireproof asbestos spray formerly used for insulating construction sites. In the fall of 1969 and spring of 1970, two major construction projects where Cafco D was used were the World Trade Center (‘WTC”) in Manhattan and the Meadow-brook Hospital (“Meadowbrook”) in Nassau County, New York.
Maiorana was employed as a sheet metal worker for a small company which performed sheet metal work on both the WTC and Meadowbrook projects. Plaintiff contends that Maiorana and the other sheet metal workers — who worked in close proximity to the asbestos sprayers — were exposed to asbestos through contact with Cafco D. See Asbestos Litig. Ill, 827 F.Supp. at 1024.
In January 1983, Maiorana was diagnosed with colon cancer. Six months later, on June 16, 1983, Maiorana died from the disease. He was 40.
*1127Plaintiff filed her original complaint on July 28, 1987, in connection with a case brought by sixteen plaintiffs on behalf of themselves and their deceased spouses against a number of manufacturers of asbestos-containing products.1 These manufacturers included USMP, the producer of Cafeo D. By way of several third-party complaints and impleaders, a number of third-party defendants were added to the litigation, including: (1) the general contractor for the WTC project, Tishman Realty & Construction (“Tish-man”); (2) the owner of the WTC, the Port Authority of New York and New Jersey (“the Port Authority”); (3) the general contractor for the Meadowbrook project, Cas-tagna & Sons (“Castagna”); and (4) the asbestos spray contractor for Meadowbrook and for the WTC interior, Mario & DiBono Plastering (“Mario & DiBono”). Id. at 1023 & n. 2.
In a series of rulings in 1991, the district court awarded summary judgment in favor of defendants, including USMP, on the grounds that the epidemiological and clinical evidence of causation were insufficient to meet the preponderance standard. See In re Joint E. & S. Dist. Asbestos Litig., 758 F.Supp. 199 (S.D.N.Y.1991) (“Asbestos Litig. I ”), reargument denied, 774 F.Supp. 113, and reconsideration denied, 774 F.Supp. 116 (1991).
On appeal, we reversed the grant of summary judgment and remanded for further proceedings, concluding that the evidence was sufficient to survive summary judgment. In re Joint E. & S. Dist. Asbestos Litig., 964 F.2d 92, 96-97 (2d Cir.1992) (‘Asbestos Litig. II ”). We found that plaintiff had presented not only epidemiological studies in support of a causal connection between- asbestos exposure and colon cancer, but also clinical evidence — in the form of Maiorana’s own medical records and personal history, which plaintiff’s experts used to exclude other possible causal factors. See id. at 96-97. We found that the statements of the plaintiffs experts, viewed in the light most favorable to plaintiff, were the “equivalent of stating that asbestos exposure more probably than not caused the colon cancer.” Id. at 97.
From January 20 to February 10,1993, the ease was tried before a jury. By the time the jury was ready to deliver its verdict, all the original direct defendants but USMP had settled. Asbestos Litig. Ill, 827 F.Supp. at 1023 n. 2. The jury found in favor of plaintiff in the amount of $4,510,000. After allocating percentages of fault among USMP and third-party defendants, the jury found USMP 50% responsible for plaintiffs damages and found three of the third-party defendants approximately equally negligent (both Tishman and Castagna were assessed to be 14% responsible; subcontractor Mario & DiBono was assessed to be 15% responsible). In addition, the jury absolved the Port Authority of any liability.
On March 10,1993, USMP moved for judgment as a matter of law, pursuant to Rule 50(b) of the Federal Rules of Civil Procedure. USMP also moved, in the alternative, for (1) a new trial on grounds of the plaintiffs testimony, the conduct of the plaintiffs counsel, alleged defects in the jury instructions, and the alleged insufficiency of the evidence, and (2) remittitur on the grounds that the award was excessive. Third-party defendants Castagna, Tishman and Mario & DiBono each moved for judgment as a matter of law with respect to the portion of the jury verdict finding it liable. In addition, Tishman and Castagna moved for common-law indemnification against Mario & DiBono.
B. The District Court’s Finding of Insufficiency
In an extensive and thoughtful opinion dated July 23, 1993, the district court granted USMP’s motion for judgment as a matter of law. In re Joint E. & S. Dist. Asbestos Litig., 827 F.Supp. 1014 (S.D.N.Y.1993) (“Asbestos Litig. III") The district court based its decision on its findings that (1) plaintiffs epidemiological evidence was insufficient to support a causal connection between asbestos exposure and colon cancer, and (2) plaintiff had failed to present affirmative clinical evi*1128dence to overcome the paucity of statistically significant epidemiological proof. Id. at 1050-51.
Epidemiology is the study of disease patterns in human populations. It “attempts to define a relationship between a disease and a factor suspected of causing it.” Brock v. Merrell Dow Pharmaceuticals, Inc., 874 F.2d 307, 311 (5th Cir.), modified on reh’g, 884 F.2d 166 (5th Cir.1989), cert. denied, 494 U.S. 1046, 110 S.Ct. 1511, 108 L.Ed.2d 646 (1990). As the district court observed, epidemiological evidence is indispensable in toxic and carcinogenic tort actions where direct proof of causation is lacking. See id. at 1026-28.
Epidemiologists speak in the statistical language of risks and probabilities. The relative risk that exposure to a given causal factor (“c”) will lead to a certain disease (“d") is expressed as a single-digit ratio, which in the context of asbestos litigation is known as the “standardized mortality ratio” (“SMR”). Id. at 1028. An SMR of 1.0 is the expected rate of contracting d in a population not influenced by c, the causal factor under investigation. An SMR of 2.0 means that d was as likely as not to have been caused by c, and an SMR greater than 2.0 means that d was more likely than not caused by c. See id. at 1027 (quoting Manko v. United States, 636 F.Supp. 1419, 1434 (W.D.Mo.1986), aff'd and remanded, 830 F.2d 831 (8th Cir.1987)).
In order for plaintiff to present a jury question on the issue of causation, the district court noted that she bore the burden of demonstrating that asbestos exposure was “more likely than not” the cause of Maiora-na’s colon, cancer. Id. at 1029-30 (citing In re Agent Orange Prod. Liab. Litig., 597 F.Supp. 740, 785 (E.D.N.Y.1984) (“Agent Orange I”), aff'd, 818 F.2d 145 (2d Cir.1987) cert. denied, 484 U.S. 1004, 108 S.Ct. 695, 98 L.Ed.2d 648 (1988)). This burden could be met either through studies conclusively establishing an SMR of more than 2.0, or through epidemiological evidence falling short of 2.0 in combination with “clinical or experimental evidence which eliminates eon-founding factors and strengthens the connection between c and d specifically in the circumstances surrounding the plaintiffs ease of d.” Id. at 1030.
The district court then evaluated the epidemiological studies in evidence, noting the SMRs reached in each study and assessing the reliability of each study in light of five “Sufficiency Criteria” derived from the work of epidemiologist A. Bradford Hill.2 These criteria include “strength and consistency of association” between c and d; the “dose-response relationship” between c and d; experimental evidence; the “plausibility” of a link between c and d; and the “coherence” between c and d. Id. at 1037-38. The meaning of each of these Sufficiency Criteria will be described in turn.
1. Strength and Consistency of Association
According to the district court, strength of association is “measured by the relative risk or the ratio of the disease rate in those with the factor to the rate in those without.” Id. at 1038 (quoting Landrigan v. Celotex Corp., 127 N.J. 404, 605 A.2d 1079, 1086 (1992)). The district court describes consistency of association as “measured by comparing the association between a purported cause and effect identified in one study with the results of other studies and with other relevant scientific knowledge.” Id.
Plaintiffs expert witnesses — Dr. Steven Markowitz and Dr. Carl Shy — testified at trial that there was a causal relationship between asbestos exposure and colon cancer. The district court, however, conducted an independent and detailed analysis of many of the epidemiological studies and concluded that there was no basis for plaintiffs experts’ conclusions. To this end, the district court noted that many of the studies yielded SMRs falling between 1.0 and 1.5, which the court deemed “statistically insignificant.” Id. at 1041-42.
A few studies adduced by plaintiff did yield SMRs that the court acknowledged were sta*1129tistically significant, including studies by Sel-ikoff and Seidman3 — which yielded SMRs of 1.62 and 1.85 respectively — and by Hilt,4 which revealed an SMR of 2.27. Id. at 1040-41. The district court criticized the methodologies employed in these studies,5 however, and found that when considered in the context of all the studies, plaintiffs evidence “establishe[d] only the conclusions that the association between exposure to asbestos and developing colon cancer is, at best, weak, and that the consistency of this purported association across the studies is, at best, poor.” Id. at 1042.. The district court concluded that the epidemiological evidence failed to satisfy the Sufficiency Criterion of strength and consistency of association, and therefore failed to contribute to the sufficiency of plaintiffs proof on causation. Id. at 1043.
2. Dose-Response Relationship
The district court described the dose-response relationship — the relationship between certain doses of c and the subsequent development of d — between asbestos and colon cancer as “erratic at best.” Id. at 1044. In his analysis, the district judge refused to credit a study cited by plaintiff suggesting a relatively high dose-response relationship,6 because the court disagreed with the assumption of that study that lung cancer rates could be used — in the absence of any direct measurement — as a substitute measure for a given population’s exposure to asbestos. Id. at 1044-45.
S. Experimental Evidence
The district court found that studies done on animals also have not established a causal relationship between asbestos and colon cancer. Id. at 1046. Although Markowitz testified that asbestos exposure did increase the rate of development of pre-cancerous polyps in animals, the district court — without directly discussing the studies cited by Markow-itz — concluded that the experimental evidence lent no support to the claim that asbestos exposure and colon cancer were causally related. Id. at 1045-46.
Jp. Plausibility
This Sufficiency Criterion asks whether it is biologically plausible, in light of the biological and chemical mechanisms involved, for exposure to c to precipitate the subsequent development of d. Id. at 1038. The district court acknowledged that asbestos is generally a carcinogen, and that a causal linkage between asbestos exposure and colon cancer is “possible.” Nonetheless, the district court — without citing any specifics — stated that plaintiffs evidence “does not support the conclusion that [the] relationship [between asbestos and colon cancer] is anything more than possible.” Id. at 1046.
5. Coherence
This criterion refers to the analysis of the instant causal factor in the context of other possible causal factors. For example, a worker’s exposure to asbestos as a possible cause of his lung cancer would be lessened on a showing that he was a heavy smoker. In meeting the coherence criterion, then, plaintiff would have to use epidemiological evidence to rule out other possible confounding factors, or she would have to employ clinical *1130or other particularistic evidence to eliminate the possibility that other confounding factors were more likely than not to have caused the disease. Id. at 1046.
The district court first observed that not only were plaintiff’s experts unable to narrow down the universe of possible confounding factors for colon cancer, but they acknowledged that asbestos exposure is not considered to be a risk factor. Id. at 1047. Indeed, as the district court noted, the rate for colon cancer in Nassau County — the New York county in which Maiorana resided — was 25.7 per 100,000 persons between 1970 and 1975, compared to 18.1 per 100,000 persons nationally during the same time period.7 This fact, the district court wrote, “raises a serious question about the various carcinogenic substances to which Maiorana and all the other residents of Nassau County have been exposed over the years and which constitute confounding factors in assessing the proximate cause of Maiorana’s cancer.” Id. at 1048.
With respect to clinical evidence, the district court observed that plaintiff’s case consisted only of a “differential” diagnosis, meaning that plaintiffs efforts attempted to exclude other confounding factors based on Maiorana’s medical records — as opposed to introducing affirmative evidence of causation such as the presence of asbestos fibers in Maiorana’s cancerous tissues. Id. Plaintiffs experts testified that because Maiorana was only 40 years old at the time of his death, had no family history of cancer, suffered from no special disease or syndrome, and did not face an abnormal risk in his diet inasmuch as it was low in fat, his colon cancer must have been caused by asbestos exposure. Id. at 1049. The district court concluded, however, that this expert testimony “failed to contribute to the sufficiency of the Plaintiffs causation proof.” Id. Although the district judge acknowledged that the presence of colon cancer in a 40-year-old man was uncommon, he nevertheless concluded that “it is neither startling nor so uncommon that it constitutes a mesothelioma-like signature disease arising only when a person of that age is exposed to asbestos.” Id. The district court further pointed out that plaintiffs latency period from his first exposure to asbestos was only 13 years, a time period that the court found was too short to be probative, in light of studies revealing a typical latency period of more than 20 years from the initial exposure to asbestos. Id. at 1049-50.
In sum, the district court found that inasmuch as plaintiffs epidemiological evidence failed, in the court’s view, to satisfy any of the Sufficiency Criteria commonly employed by epidemiologists, this evidence was insufficient to support the general proposition that asbestos exposure causes colon cancer. Id. at 1050. The district court criticized plaintiffs experts’ conclusions as “masquerading behind the guise of sound science.” Id. Given plaintiffs failure to show that any asbestos fibers were found in Maiorana’s cancerous tissues, the district court found that the sum total of plaintiffs evidence did not justify the jury’s finding of causation, and that “the jury’s finding could only have been the result of sheer surmise and conjecture.” Id. at 1051 (quoting Samuels v. Air Transport Local 504, 992 F.2d 12, 14 (2d Cir.1993)). Therefore, concluded the district court, entry of judgment as a matter of law in favor of defendants was warranted.
C. The District Court’s Rulings on the Post-Trial Motions
Finally, the district court also ruled on the alternative motions of USMP and third-party defendants, reasoning that “[wjere the Second Circuit to reverse the entry of judgment as a matter of law on appeal, the Court of Appeals could use the following conclusions in shaping its order in lieu of simply reinstating the jury verdict.” Id. at 1051. To this end, the district court (1) denied USMP’s motion for a new trial, rejecting USMP’s claims of prejudice, or in the alternative, erroneous jury instructions regarding whether the third-party defendants constituted a superseding cause; (2) denied motions for judgment as a matter of law by third-party *1131defendants Castagna, Tishman and Mario & DiBono, who argued that there was insufficient evidence that they had received notice of the dangers of asbestos or failed to act reasonably; (3) conditionally granted defendant’s motion for remittitur and reduced the jury verdict by $1.2 million, decreasing the total award to $3.31 million; and (4) denied Tishman’s and Castagna’s motions for common-law indemnification from Mario & DiBo-no.
II. The SUFFICIENCY of Plaintiff’s PROOF of Causation
The central issue in this appeal is whether a district judge’s expanded role in assessing the admissibility of scientific evidence in federal trials, as articulated by the Supreme Court in Daubert, extends to consideration of the sufficiency of scientific evidence already admitted. In the present case, the sufficiency inquiry bears on the factual issue of causation — whether the body of plaintiffs evidence was sufficient to persuade a rational jury that Maiorana’s exposure to asbestos more likely than not caused his colon cancer. Causation in toxic torts normally .comprises two separate inquiries: whether the epidemiological or other scientific evidence establishes a causal link between c (asbestos exposure) and d (colon cancer), and whether plaintiff is within the class of persons to which inferences from the general causation evidence should be applied. See In re Agent Orange Product Liability Litigation, 611 F.Supp. 1223, 1261-62 (E.D.N.Y. 1985) (“Agent Orange II”), aff'd, 818 F.2d 187 (2d Cir.1987), cert. denied, 487 U.S. 1234, 108 S.Ct. 2898, 101 L.Ed.2d 932 (1988).
For the reasons stated below, we hold that the district court overstepped the boundaries set forth in Daubert. It impermissibly crossed the line from assessing evidentiary reliability to usurping the role of the jury. Accordingly, we reverse the district court’s entry of judgment as a matter of law with respect to the jury verdict in favor of plaintiff.
A. Standard for Judgment as a Matter of Law
It is well established that a district court should not grant a motion for judgment as a matter of law except on a showing of “such a complete absence of evidence supporting the verdict that the jury’s findings could only have been the result of sheer surmise and conjecture.” Samuels, 992 F.2d at 14 (internal quotation omitted). In considering a motion for judgment as a matter of law, the district court “must view the evidence in a light most favorable to the nonmovant and grant that party every reasonable inference that the jury might have drawn in its favor.” Purgess v. Sharrock, 33 F.3d 134, 140 (2d Cir.1994). In this regard, the district court is not to “assess the weight of conflicting evidence, pass on the credibility of the witnesses, or substitute its judgment for that of the jury.” Mattivi v. South African Marine Corp., “Huguenot,” 618 F.2d 163, 167 (2d Cir.1980).
In reviewing the district court’s grant of USMP’s motion for judgment as a matter of law, we apply the same standard as the district court. Arkin v. Gittleson, 32 F.3d 658, 664 (2d Cir.1994). Thus, we must determine whether, drawing all reasonable inferences regarding the weight of the evidence and the credibility of witnesses in favor of plaintiff, a reasonable jury could only have found for the defendants. Samuels, 992 F.2d at 14.
As the district court opinion notes, the Daubert Court did point to the granting of summary judgment and judgment as a matter of law as “appropriate safeguards” where a ease built on shaky scientific evidence is insufficient to go before a jury. See Daubert, — U.S. at -, 113 S.Ct. at 2798. In so doing, however, Daubert did not alter the standards for judgment as a matter of law in scientific cases: Justice Blackmun, writing for the court, referred to entry of summary judgment and judgment as a matter of law as “conventional devices,” id., and he spoke about the possibilities afforded by those remedies. For reasons described more fully below, the standards by which we review the district court’s entry of judgment as a matter of law in this case — as set forth above — are in no way modified by Daubert.
*1132 B. Admissibility, Sufficiency, and the Daubert case
The “admissibility” and “sufficiency” of scientific evidence necessitate different inquiries and involve different stakes. Admissibility entails a threshold inquiry over whether a certain piece of evidence ought to be admitted at trial. The Daubert opinion was primarily about admissibility. It focused on district courts’ role in evaluating the methodology and the applicability of contested scientific evidence in admissibility decisions. See — U.S. at-, 113 S.Ct. at 2796-98.
This case is about sufficiency, not admissibility. A sufficiency inquiry, which asks whether the collective weight of a litigant’s evidence is adequate to present a jury question, lies further down the litigational road. In the case at bar the admissibility of the epidemiological studies was not at issue: the studies had already been admitted at trial. See Asbestos Litig. Ill, 827 F.Supp. at 1025 (“[Tjhere is no question in this case that the testimony of both sides’ experts — all of whom possess impressive credentials — is admissible.”). The issue before the district court in this case, then, was whether the epidemiological and clinical data already in ‘evidence was sufficient to justify the jury’s verdict finding causation.
The Daubert Court significantly changed the standards governing the admissibility of scientific evidence by expanding district courts’ discretions to evaluate the reliability and relevance of contested evidence. — U.S. at -, 113 S.Ct. at 2799. USMP argues that Daubert ushered in similar changes with respect to sufficiency decisions, claiming that “under Daubert, admissibility and sufficiency determinations are functionally equivalent.” Appellee Br. at 23. We find no textual support in the Daubert opinion for USMP’s interpretation, and we conclude that Daubert left the traditional sufficiency standard intact.
The issue of sufficiency was not squarely before the Daubert Court, but Justice Blaek-mun’s opinion did briefly address the issue. Once certain pieces of scientific evidence pass the admissibility threshold, Justice Blackmun wrote, the “appropriate” means of challenging those which appear shaky or unreliable include the “traditional” devices of “[vjigorous cross-examination, presentation of contrary evidence, and careful instruction on the burden of proof.” — U.S. at -, 113 S.Ct. at 2798. In addition, as noted above, the Daubert Court reiterated the possibility of a district court’s granting summary judgment or judgment as a matter of law in cases where “the scintilla of evidence presented supporting a position is insufficient to allow a reasonable juror to conclude that the position more likely than not is true.” Id. The passages in Daubert regarding sufficiency sounded a tone of optimism about juries’ abilities to sort out the veracity of competing scientific evidence. The Court rejected one litigant’s claim that abandoning the “general acceptance” requirement of admissibility, Frye v. United States, 293 F. 1013, 1014 (App.D.C.1923),8 would “result in a ‘free-for-all’ in which befuddled juries are confounded by absurd and irrational pseudoscientific assertions.” Daubert, — U.S. at -, 113 S.Ct. at 2798. By the same token, the Court characterized that litigant as “overly pessimistic about the capabilities of the jury, and of the adversary system generally.” Id.
As illustrations of appropriate uses of summary judgment or judgment as a matter of law, the Daubert Court cited two Court of Appeals opinions mandating, on grounds of insufficiency, the removal from the jury of cases grounded in epidemiological evidence. In the first case, Turpin v. Merrell Dow Pharmaceuticals, Inc., 959 F.2d 1349 (6th Cir.), cert. denied, — U.S. -, 113 S.Ct. 84, 121 L.Ed.2d 47 (1992), the Court of Appeals found that experimental evidence involving animal studies was insufficient to permit a reasonable jury to find that B endec-tin — an anti-nausea prescription for morning *1133sickness — caused plaintiff’s birth defects. Unlike Arlene Maiorana, the plaintiffs in Turpin did not present supportive epidemiological evidence in addition to their animal studies, whereas their adversaries introduced 35 epidemiological studies demonstrating no causal link. Id. at 1353. We find nothing in Turpin that suggests an alteration of the traditional standard for assessing the sufficiency of scientific evidence.
In the second case, Brock, the Court of Appeals found that plaintiffs epidemiological evidence was insufficient to support a jury verdict finding that Bendectin caused plaintiffs child’s birth defects; thus the court granted judgment as a matter of law in favor of defendants. The basis for the court’s holding was that the risk ratios yielded by the plaintiffs epidemiological studies — upon adjustment by confidence intervals — included the ratio of 1.0, which would be the standard expected rate of birth defects in a population not exposed to Bendectin ingestion. Brock, 874 F.2d at 312-13. Our reading of Brock is that it applied the traditional sufficiency standard.
Although we conclude that Daubert did not alter the traditional sufficiency standard — contrary to the position urged by USMP — we acknowledge that sufficiency poses unique difficulties for trial courts in toxic or carcinogenic tort cases, such as the one before us, which hinge on competing interpretations of epidemiological evidence. By its nature, epidemiology is ill-suited to lead a factfinder toward definitive answers, dealing as it does in statistical probabilities and the continual possibility of confounding causal factors. See Brock, 874 F.2d at 311 (“One difficulty with epidemiological studies is that often several factors can cause the same disease.”). In light of the inherent uncertainty shrouding issues of probabilistic causation, the decision of a district court on whether plaintiffs epidemiological evidence is sufficient to get to the jury should be guided by the well-established standards governing judgment as a matter of law — whether, viewed in the light most favorable to the nonmoving party, “the evidence is such that, without weighing the credibility of the witnesses or otherwise considering the weight of the evidence, there can be but one conclusion as the verdict that reasonable [jurors] could have reached.” Samuels, 992 F.2d at 14 (quoting Simblest v. Maynard, 427 F.2d 1, 4 (2d Cir.1970)).
Applied to epidemiological studies, the question is not whether there is some dispute about the validity or force of a given study, but rather, whether it would be unreasonable for a rational jury to rely on that study to find causation by a preponderance of the evidence. In addition, multiple epidemiological studies cannot be evaluated in isolation from each other. Unlike admissibility assessments, which involve decisions about individual pieces of evidence, sufficiency assessments entail a review of the sum total of a plaintiffs evidence. It is with these general principles in mind that we turn to a review of the district court’s analysis of plaintiffs epidemiological and clinical evidence in support of causation in the case at bar.
C. Review of the District Court’s Analysis of the Scientific Evidence
While the district court’s analysis was certainly thorough and well-documented, it was rife with independent assessments of witnesses’ conclusions and comparative credi-bilities, often in a manner that appears to us to stretch the above-cited passages in Dau-bert beyond their limit. We believe that the district court, in many instances, did engage in the proscribed practices of “assess[ing] the weight of conflicting evidence, pass[ing] on the credibility of the witnesses [and] substituting] its judgment for that of the jury.” Smith v. Lightning Bolt Prods., Inc., 861 F.2d 363, 367 (2d Cir.1988). What follows is a non-exhaustive list of those instances in which we hold the district court to have been in error.
1. The district court disregarded Mar-kowitz’s testimony that asbestos fibers have been found in the cancerous tissues of some asbestos workers.. Markowitz based his opinion on Ehrlich’s finding that, in a study of a population sample comprising 44 asbestos workers who had developed colon cancer, asbestos fibers were present in the cancerous *1134tissue of 14 of them.9 The district court deemed this study “irrelevant” on the grounds that “there was no finding in the pathology reports at issue here that there were asbestos fibers in Maiorana’s tumorous tissues.” Asbestos Litig. Ill, 827 F.Supp. at 1039. In so doing, the district court conflated the general and clinical aspects of the causation issue. Plaintiff relied on Ehrlich’s study to bolster the general contention that there is a causal link between asbestos exposure and colon cancer. Whether plaintiff established that link as a general proposition, or whether Ehrlich’s study was reliable in support of that proposition, had nothing to do with the clinical specifics of Maiorana’s case.
2. The district court discredited Markow-itz’s testimony that a number of studies listing SMRs for asbestos/colon cancer ranging from 1.14 to 1.47 are statistically significant when taken together. Asserting that an SMR of less than 1.50 is “statistically insignificant,” id. at 1041, the district court stated, “no matter how many studies yield a positive but statistically insignificant SMR for colo-rectal cancer, the results remain statistically insignificant. Just as adding a series of zeros together yields yet another zero as the product, adding a series of positive but statistically insignificant SMRs together does not produce a statistically significant pattern.” Id. at 1042.
The district court, however, cited no authority for the bold assertion that SMRs of less than 1.50 are statistically insignificant and cannot be relied upon by a jury to support a finding of causation. Although perhaps a floor can be set as a matter of law, we are reluctant to adopt such an approach. We believe that it would be far preferable for the district court to instruct the jury on statistical significance and then let the jury decide whether many studies over the 1.0 mark have any significance in combination.
3. The district court erred in disregarding the studies by Selikoff and Seidman— revealing SMRs of 1.6210 and 1.8511 for asbestos exposure and colon cancer — and by Hilt, whose study yielded an SMR of 2.27 for colon cancer in plant workers exposed to asbestos.12 These SMRs exceeded the district court’s own threshold for statistical significance, yet the court concluded that they “failed to contribute to the sufficiency of [plaintiffs] proof on the issue of causation.” Id. at 1043.
The district court did not specify its basis for disregarding the above three studies. It criticized the methodology of each study, plus it suggested that the causal connection between asbestos exposure and colon cancer must “be supported by all relevant epidemiological studies.” Id. at 1041. In our view, neither assessment provides a basis for the court’s wholesale rejection of the studies.
The district court cites to a published attack on SelikofPs and Seidman’s methodology.13 This criticism might have undercut the evidentiary weight of those studies, but it did not leave the jury without a basis for concluding that the studies tended to prove plaintiffs claim. Similarly, the district court’s criticism of Hilt’s study for its allegedly limited sampling size provided no basis for holding — once the study was already deemed admissible — that the jury could not, as a matter of law, ascribe any probative significance to that study.
By the same token, to the extent the district court found that the conclusions of the above three studies were less persuasive than the contrary conclusions of defendant’s studies,14 that finding would have constituted an inappropriate weighing of credibilities. *1135The district court was bound to consider the evidence in the light most favorable to plaintiff and to grant plaintiff “every reasonable inference that the jury might have drawn in its favor,” Purgess v. Shamrock, 33 F.3d at 140. To this end, it was perfectly reasonable for a jury to credit the collective probative weight of the conclusions reached by Selikoff, Seidman, Hilt, and the other epidemiological evidence.
Finally, we note that nowhere in its opinion does the district court mention Albiris 1990 study, which USMP’s expert witness, Gaensler, acknowledged yielded an SMR of 1.5 — a mark that comports with the district court’s own test for statistical significance.
4. In our estimation, the district court placed too much emphasis on certain studies, particularly those by Gaensler15 and Weiss.16 The court linked its finding that plaintiffs case was insufficient as a matter of law to the failure of plaintiffs experts to rebut studies by Gaensler and Weiss:
The Plaintiffs experts acknowledged without refuting either the conclusions derived by the various epidemiologists used in Weiss and Gaensler’s surveys or the conclusions of Weiss and Gaensler. In the face of this substantial and notably consistent body of scientific evidence, the opinions of Markowitz and Shy asserting a causal connection between asbestos and Maiorana’s colon cancer were nothing more than ‘sheer surmise and conjecture,’ masquerading behind the guise of sound science.
Asbestos Litig. Ill, 827 F.Supp. at 1050 (internal citation omitted). Further, the district court opinion relied especially heavily on the Weiss report, citing to it with approval in 12 separate footnotes. See id. at nn. 18, 43, 45, 46, 54, 58, 62, 64, 66, 69, 78, 79. In so doing, the district court once again failed to fulfill the requirement of considering the epidemiological evidence in the light most favorable to plaintiff and substituted its own scientific conclusions for those of the expert witnesses. The court erred in relying so heavily on certain studies submitted by defendants, and in insisting that plaintiff rebut the conclusions drawn in those studies.17 For the district court to seize on the putative flaws of studies favorable to plaintiff, and then to privilege certain studies favorable to defendant, was impermissibly to place a thumb on defendant’s side of the scale and to encroach on the jury’s prerogative to weigh the relative merits and credibilities of competing studies. In this regard, we note that even the Weiss study was flawed: Defendant’s own expert witness, Gaensler, admitted that Weiss’s study did not employ a sound statistical approach, and agreed that the approach was “not quite kosher.” J.A. 886. Thus, to the extent that none of the studies is flawless or dispositive, their relative merits seems to us to be a classic question for the jury. Trial courts should not arrogate the jury’s role in “evaluating the evidence and the credibility of expert witnesses” by “simply cho[o]s[ing] sides in [the] battle of the experts.” Christophersen v. Allied-Signal Corp., 902 F.2d 362, 366 (5th Cir.1990).
5. Conspicuously absent from the district court opinion was any mention of federal government agency reports that were before the jury.18 In these reports, both the Occupational Safety and Health Administration (“OSHA”) and the Environmental Protection Agency (“the EPA”) concluded, upon their review of available epidemiological studies, that a strong causal link does exist between asbestos exposure and gastrointestinal cancer, including colon cancer.19 A rational jury *1136could have decided that these agencies’ conclusions bolstered Markowitz’s and Shy’s expert opinions that the epidemiological evidence does establish a causal connection between asbestos exposure and colon cancer. For the district judge to supersede the opinions of the expert witnesses with his own lay judgment raises some concerns; for the court to omit any consideration of agency reports backing up the claims of plaintiffs experts and supporting the jury verdict is, in our view, especially troubling.20
6. In our view, the district court opinion impermissibly disregarded several statements we made in our 1992 opinion reversing the court’s granting of summary judgment in favor of defendants.
First, the district court stated the following: “[Djespite the length and detail of the testimony given by the Plaintiffs expert witnesses and the self-assured manner in which they stated the[ir] conclusory opinions, ‘there [was] such a complete absence of evidence supporting the verdict that the jury’s finding could only have been the result of sheer surmise and conjecture_’” Asbestos Li-tig. Ill, 827 F.Supp. at 1051 (citation omitted). This assertion seems at variance with our previous statement that
[Plaintiffs] experts concluded that asbestos exposure was “a significant factor” and “a proximate cause, and a substantial factor” in causing the colon cancer. Granting plaintiff all favorable inferences, these statements are the equivalent of stating that asbestos exposure more probably than not caused the colon cancer. Medical experts need not invoke technical legal phrases in order to convey their medical opinions.”
Asbestos Litig. II, 964 F.2d at 97.
Second, the district court again attacked the testimony and credentials of Dr. Roth-man21 — the physician who treated Maiora-na — in spite of our prior holding that the district court “unduly minimized the expertise of Dr. Rothman, who is a specialist in internal medicine.” Id. We note that we also criticized the district court’s rejection of Rothman at the summary judgment stage as “overly harsh.” Id.
Third, the district court stated that plaintiff’s case was devoid of “affirmative clinical evidence.” 827 F.Supp. at 1050; see also id. at 1025 (characterizing plaintiffs evidence as “exclusively scientific in nature”). This characterization contradicted our prior conclusion that Maiorana’s own medical records and personal history were adequate to serve as a clinical basis for a conclusion that asbestos exposure constituted a significant causal factor in Maiorana’s colon cancer.22
*11377. The district court placed what we believe was undue emphasis on defendants’ use of animal studies and on plaintiff’s comparative lack of such evidence. The lack of experimental evidence supplied no basis for removing the case from the jury, especially inasmuch as plaintiff did adduce epidemiological evidence.
8. In rejecting plaintiffs experts’ differential diagnosis, the district court gave little weight to Maiorana’s relatively young age of death. While conceding it was “uncommon for a 40-year-old man to develop colon cancer,” 827 F.Supp. at 1049, the court emphasized the absence of asbestos fibers in Maior-ana’s colon cancer tissues and the typical latency period of more than 20 years for colon cancer, compared to only 13 years for Maiorana. Id. The district court’s conclusions regarding latency periods, however, ignored Markowitz’s testimony on the study by Seidman which indicated an increased incidence of gastrointestinal cancer among individuals exposed to asbestos 10 years previously. That a latency period of 13 years was too short appeared to constitute an independent medical conclusion by the district court.
D. Conclusion
For the above reasons, we hold that the district court erred in ruling that plaintiff presented insufficient epidemiological and clinical evidence to support the jury’s verdict finding causation. In our view, the district court impermissibly made a number of independent scientific conclusions — without granting plaintiff the requisite favorable inferences — in a manner not authorized by Daubert. As Chief Justice Rehnquist has written, the law does not “impose[] on [judges] either the obligation or the authority to become amateur scientists.” Daubert, — U.S. at-, 113 S.Ct. at 2800 (Rehnquist, C.J., concurring in part and dissenting in part).
III. The Post-TRial Motions
Having reversed the district court’s entry of judgment as a matter of law, we now turn to a review of the appeals filed by the third-party defendants/cross-appellants. First, third-party defendants Castagna and Tish-man challenge the district court’s denial of their claim that there was insufficient evidence to support the jury’s finding that they had constructive notice of the dangers of asbestos. Second, third-party defendant Mario & DiBono appeals from the portion of the jury verdict finding it 15 percent hable for plaintiffs injuries, on the ground that USMP failed to make a prima facie case against it. Third, Tishman and Castagna both appeal from the district court’s denial of their motions for indemnification against Mario & DiBono.23
For the reasons stated below, we affirm the jury verdict and the rulings of the district court with respect to each claim brought by the third-party defendants. Further, in view of the district court’s conditional order of a remittitur, we remand the case to the district court so that plaintiff will have the opportunity to decide whether to accept the remittitur, or submit to a new trial.
A. The Sufficiency Challenge by Castagna and Tishman
Castagna and Tishman — the primary contractors on the two construction projects in which Maiorana worked — argued following the trial and on appeal that there was insufficient evidence to support the jury’s finding that they were on constructive notice of the dangers of asbestos. Without such notice, they argue, the protections for workers of § 200 of New York Labor Law do not apply to any negligence on their part in the handling of USMP’s asbestos-containing products. We agree, however, "with the district court’s finding that there was sufficient evidence in the record to support the jury’s finding that Castagna and Tishman were adequately put on constructive notice.
Constructive notice is imputed to an employer who should reasonably have known *1138of a potential hazard; it triggers on the employer’s part the duty to “exercise reasonable care to inform and protect its employees.” Gallose v. Long Island R.R., 878 F.2d 80, 84-85 (2d Cir.1989). There was ample evidence in the record from which a rational jury could have concluded that Castagna and Tishman had actual or at least constructive notice of hazards associated with asbestos dust. This evidence included USMP’s placing of caution labels on containers comprising Cafco D and other asbestos-containing products24; regulatory measures for asbestos promulgated by state authorities25 and by OSHA26; Castagna’s and Tishman’s presence at safety meetings in which the dangers of asbestos were discussed; and numerous news reports on the hazards posed by asbestos.27
Further evidencing constructive notice on Castagna’s part was a letter from HRH Construction Corp. written to the Nassau County Department of Public Works on October 2, 1969, in which HRH complained that the general contractor was not cleaning up the dust from the asbestos spraying. With respect to Tishman, there was evidence in the record that Mario & DiBono complained that Tishman failed to supply containers to carry away cleaned-up dust, and that Tishman failed to arrange for worker safety meetings. We hold that this evidence, taken together, provided an ample basis for a rational jury to conclude that Castagna and Tishman had notice of the hazards of asbestos dust.
B. The Sufficiency Challenge by Mario & DiBono
Third-party defendant Mario & Di-Bono, the subcontractor responsible for the actual spraying of the asbestos, argues on appeal that the district court erred in declining to vacate the portion of the jury verdict finding Mario & DiBono partially hable. We find that Mario & DiBono was clearly on notice — not only was the same factual evidence which demonstrated constructive notice on the parts of Castagna and Tishman introduced against Mario & DiBono, but USMP also supplied evidence that Mario & DiBono had actual notice, including the warning labels on the Cafco D containers, and a 1968 Sales and Application Manual identifying various safety measures.
Regarding Mario & DiBono’s alleged negligence, the issue on appeal is whether there was a triable question of fact. We believe that there was. As the district court notes, in spite of Mario & DiBono’s contractual obligation to clean up the asbestos-containing debris remaining after its work, conditions at both the World Trade Center and Meadowbrook sites were such that both Cas-tagna and Tishman registered complaints against Mario & DiBono for inadequately preventing loose dust from blowing into surrounding areas. Hence, there was sufficient evidence of Mario & DiBono’s negligence to submit to the jury.
C. Castagna’s and Tishman’s Indemnification Claim
Castagna and Tishman appeal from the district court’s denial of their motions for common-law indemnification from Mario & DiBono. On this issue, New York law provides that where the cause of an accident is shared and contribution is permissible, the appropriate result is contribution and not indemnity. See Kelly v. Diesel Constr. Div. of Carl A. Morse, Inc., 35 N.Y.2d 1, 6-7, 358 N.Y.S.2d 685, 315 N.E.2d 751 (1974). The jury apportioned, in relatively equal parts, the comparative faults of Castagna, Tishman *1139and Mario & DiBono. This result bars Cas-tagna’s and Tishman’s claims for indemnity.
D. The Remittitur
Finally, we address briefly the district court’s conditional granting of USMP’s motion for remittitur and the court’s reduction of the jury verdict. Remittitur may be defined as “the process by which a court compels a plaintiff to choose between reduction of an excessive verdict and a new trial.” Earl v. Bouchard Transp. Co., 917 F.2d 1320, 1328 (2d Cir.1990). Inasmuch as the district court also granted judgment as a matter of law in favor of defendants, plaintiff has not yet been confronted with the decision whether to accept the reduced verdict or to submit to a new trial. Plaintiff must be afforded the opportunity to make this choice. See Phelan v. Local 305, 973 F.2d 1050, 1064 (2d Cir.1992) (“[A] court should not decrease damages based on its belief that they are excessive without affording the option of a new trial.”), cert. denied, — U.S. -, 113 S.Ct. 1415, 122 L.Ed.2d 785 (1993).
Accordingly, we remand this case to the district court, at which point plaintiff must decide whether to accept the reduced verdict as a condition of the district court’s denial of USMP’s motion for a new trial. See Tingley Sys. v. Norse Sys, 49 F.3d 93 (2d Cir.1995). The reduced verdict will not be ripe for appeal unless plaintiff rejects the reduced verdict and completes a new trial. Earl, 917 F.2d at 1328.
IV. Conclusion
By way of summary:
1.On the issue of causation, we reverse the district court’s entry of judgment as a matter of law, and we reinstate the jury verdict in favor of plaintiff. The district court erred in failing to draw all reasonable inferences in favor of plaintiff; in rendering independent assessments of the epidemiological evidence far beyond the role authorized by Daubert; in rejecting all epidemiological studies that yielded an SMR below the unexplained floor of 1.50; in rejecting several of plaintiffs other studies for what appear to be reasons of credibility; and in generally encroaching upon the factfinding role of the jury.
2. With regard to the cross-appeals of the third-party defendants, we affirm the district court’s denial of Castagna’s and Tishman’s insufficiency claims; affirm the jury verdict finding Mario & DiBono partially liable; and affirm the district court’s denial of Castag-na’s and Tishman’s motions for indemnification against Mario & DiBono.
3. In view of the conditional order for a remittitur, we remand to the district court to provide plaintiff with the requisite opportunity to decide whether to accept the reduced verdict or submit to a new trial.
The district court’s entry of judgment as a matter of law in favor of defendants is reversed, and the jury’s verdict — as modified by the district court — is reinstated. The court’s disposition of those motions which are appealed by the third-party defendants are affirmed. The case is remanded for further proceedings consistent with this opinion.
. In May 1988, plaintiff's claims were separated from those of the other plaintiffs and filed as an independent case.
. A. Bradford Hill, The Environment and Disease: Association or Carnation, 58 Proc. Royal Soc'y Med. 295, 295-300 (1965).
. Irving J. Selikoff, et al., Mortality Experience of Insulation Workers in the United States and Canada, 1943-1976, 330 Annals N.Y. Acad. Sci. 91, 108 (1979); Herbert Seidman, et al., Mortality Experience of Amosite Asbestos Factory Workers: Dose-Response Relationships 5 to 40 Years After Onset of Short-Term Work Exposure, 10 Am.J.Indus.Med 479 (1986).
. Bjorn Hilt et al., Asbestos Exposure, Smoking Habits, and Cancer Incidence Among Production and Maintenance Workers in an Electrochemical Plant, 8 AmJ.Indus.Med. 565, 571 (1985).
. The court pointed out that another scientist had attacked Selikoff’s and Seidman's use of the "best evidence” approach' — using data derived from sources other than death certificates to compute SMRs — because that approach “introduces an erroneous bias into the calculation of SMRs and results in their being unjustifiably inflated.” Id. at 1041, n. 32 (citing Robert W. Morgan et al.. Asbestos and Gastrointestinal Cancer: A Review of the Literature, 143 W.J.Med. 60, 63-64 (1985)). Similarly, the district court asserted that Hilt's study “can be seriously questioned” for its allegedly limited sample size. Id. at 1041 n. 36.
.Howard Frumkin & Jesse Berlin, Asbestos Exposure and Gastrointestinal Malignancy Review and Meta-Analysis, 14 Am.J.Indus.Med. 79 (1988).
. N.J. Def’t of Envtl. Protection, Office of Cancer and Toxic Substance Res., Trends in Cancer Mortality in the New Jersey/New York/Philadelphia Region, 1950-1975: Data Summary (1981).
. The Frye rule states that expert opinion based on a scientific technique is inadmissible unless “the thing from which the deduction is made [is] sufficiently established to have gained general acceptance in the particular field in which it belongs.” 293 F. at 1014. The Daubert Court overruled this standard, holding that " 'general acceptance’ is not a necessary precondition to the admissibility of scientific evidence." - U.S. at-, 113 S.Ct. at 2799.
. Albert Ehrlich et al., Asbestos Bodies in Carcinoma of Colon in an Insulation Worker With Asbestosis, 254 J.Am.Med.Ass'n 2932 (1985).
. See Selikoff, supra note 3, at 108.
. See Seidman et al., supra note 3, at 483.
. See Hilt, supra note 4, at 571.
. See supra note 5 and accompanying text.
.The district court did not explicitly state that it was finding as such, but it did assert that the studies with the largest sampling sizes were "statistically the most powerful and compelling,” id. at 1042, and it stated that studies yielding lower SMRs than those derived by Selikoff, Seidman and Hilt "seriously undermined [plaintiff's experts'] claim that there is a causal linkage between asbestos and colorectal cancer." Id. at 1043.
. The district court opinion describes the work of Gaensler — who served as expert witness for USMP — without supplying any citations to Gaen-sler’s studies.
. William Weiss, Asbestos and Colorectal Cancer, 99 Gastroenterology 876, 881 (1990).
. The district court placed the same burden on plaintiff regarding the Garabrant report, pointing out that plaintiff's experts did not dispute Gara-brant's finding of an SMR of 1.14. David H. Garabrant et al., Asbestos and Colon Cancer: Lack of Association in Large Case-Control Study, 135 AmJ.Epidemiology 843, 846 (1992).
. Excerpts from these studies were read into the record at trial. See Trial Tr. at 942-48, in Addendum to Br. for Pl.-Appellant.
. See Final Rules on Occupational Exposure to Asbestos, Tremolite, Anthrophyllite, and Actinol-ite; Final Rules, 51 Fed.Reg. 22612, 22620 (June 20, 1986) ("In summary, 12 different epidemio*1136logical studies on a variety of occupational cohorts exposed to asbestos have found express mortality from gastrointestinal cancers; of these, 7 studies found statistically significant excesses. OSHA believes that these findings constitute substantial evidence of an association between asbestos exposure and a risk of incurring gastrointestinal cancer.”); Final Rule on Asbestos; Manufacture, Importation, Processing, and Distribution in Commerce Prohibitions, 54 Fed.Reg. 29461, 29469 (July 12, 1989) ("[A]fter weighing available information, EPA believes that there is evidence of a strong causal relationship between asbestos exposure and gastrointestinal cancer excess.").
.That both the EPA and OSHA reports focused on the broader area of gastrointestinal cancers (which include cancers of the esophagus, stomach, rectum as well as the colon) — rather than specifically on colon cancer — is no argument to disregard them. Drawing all reasonable inferences in favor of plaintiff, it would be fair to assume that the jury found the agency’s conclusions to be as applicable to colon cancer as to cancers of other parts of the digestive tract.
. See Asbestos Litig. III, 827 F.Supp. at 1049 (“[T]he Plaintiff called Doctor Nathan Rothman ('Rothman'), a specialist in internal medicine who treated Maiorana. Rothman, an expert in pulmonary medicine and lung disease, undertook a differential diagnosis of Maiorana, but because he acknowledged that he was not an expert in asbestos-related diseases, was not familiar with either the epidemiological or experimental literature on the subject, did not know the latency period for colon cancer, and had never undertaken any research regarding colon cancer or diseases of the colon, he was not qualified to draw the conclusion from his differential diagnosis that asbestos exposure was the cause of Maiora-na's colon cancer.”). Rothman’s testimony, however, focused on observations about Maiora-na and was offered in support of plaintiff's efforts to eliminate confounding factors. To this end, there was no need for Rothman to be an expert in asbestos-related diseases.
. See Asbestos Litig. II, 964 F.2d at 97 (“It is important to point out that Mr. Maiorana’s own medical records and personal history were the clinical evidence upon which plaintiff's expert *1137witnesses based their opinions that asbestos exposure. was a significant factor in causing his colon cancer.”) (emphasis in original).
. USMP does not appeal from the district court's denial of USMP's alternative motion for a new trial.
. These labels included the following warning: “CAUTION. This product contains asbestos. Inhalation of asbestos dust over long periods may be harmful.”
. State materials admitted into evidence included the Recommended Code of Practices for Application of Sprayed Fireproofing Materials and the New York Department of Labor Code Rule 12-27(C), which imposed a threshold level for asbestos on construction sites in New York. See Asbestos Litig. Ill, 827 F.Supp. at 1056.
. USMP introduced into evidence OSHA’s Occupational Safety and Health Standards for asbestos of 1972.
. News reports on the link between asbestos and cancer have been held to be sufficient to raise a triable issue of fact over whether a defendant had constructive notice of the dangers of asbestos. See Barnett v. City of Yonkers, 731 F.Supp. 594, 600 (S.D.N.Y.1990).