[243]*243GOETTEL, District Judge:
V olkswagenwerk Aktiengesellschaft (“VWAG”) appeals from a judgment entered against it on May 9,1979 for damages arising out of an automobile collision between plaintiffs-appellees, Turi and Frank Caiazzo, who were driving a minibus manufactured by appellant, and defendant James Valentine, who has not joined in this appeal.1
The Caiazzos’ claim against VWAG arises out of an area of tort law that has come to be. known as the “crashworthiness” or “second collision” doctrine.2 Under this theory of liability, the claimant does not allege that any defect in the automobile caused the accident; rather, the allegation is that the claimant’s injuries were more severe than what they would have been had the ear been properly designed. Appellant raises several significant issues connected with this theory of liability. This being a diversity action, we must apply New York law in resolving these issues, Erie Railroad v. Tompkins, 304 U.S. 64, 58 S.Ct. 817, 82 L.Ed. 1188 (1938), and must join other federal courts in the difficult task of predicting a state court’s interpretation in this uncertain area of the law. For the reasons set forth below, we reverse in part the judgment of the district court and remand for a new trial on damages.
I. The Factual Background
On January 17, 1972, Turi Caiazzo, accompanied by her husband, Frank Caiazzo, was driving a 1965 Volkswagen minibus westbound in the center lane of the Long Island Expressway. Neither was wearing the lap seat belt, which was anchored to the floor of the van. Defendant Valentine, proceeding at a speed estimated to be between 50 and 65 miles per hour, crashed into the rear of the Caiazzos’ van causing it to turn completely over and reverse direction. The van came to a rest facing east on the right shoulder of the expressway approximately 150 feet west of the point of impact. Both Turi and Frank Caiazzo were ejected from the van and were apparently knocked unconscious.3 Frank Caiazzo regained consciousness 100 feet west of the van on the shoulder of the expressway.4 He found Turi on the center lane of the expressway opposite the van. Neither of the Caiazzos has a detailed recollection of the accident.
Turi Caiazzo, who had a life expectancy of about 50 years at the date of trial, suffered a severe cerebral concussion, hemato-ma over the occipital region of her head, total accident amnesia, a severe fracture of her right shoulder blade, various cuts and scrapes, a myocardial contusion, and frac[244]*244tures to both ankles. (The ankle injuries were the most serious of these injuries.) As a result of these injuries, Turi was confined to a wheelchair for about two months and for some time thereafter had to use crutches. Her ankles are so sensitive that she must avoid anyone’s brushing against her foot. In addition, traumatic arthritis in her right ankle has so limited the function of the ankle that she cannot put on pants.
Frank Caiazzo’s injuries, although severe, were less extensive than his wife’s. He sustained a compression fracture of the first lumbar vertebra and a cerebral concussion. Although his back hurts him occasionally when he plays sports, he made no claim of loss of earnings or for any diminution in earning capacity. His life expectancy at the time of trial was about forty-eight years.
In 1973 the Caiazzos instituted an action against the operator of the other vehicle, Valentine, and against VWAG, the manufacturer of the Caiazzos’ van, for aggravation of injuries under a “second collision” theory. See Bohn v. Triumph Corp., 33 N.Y.2d 151, 305 N.E.2d 769, 350 N.Y.S.2d 644 (1973). The Caiazzos alleged that their injuries were aggravated by a defective door latch assembly, which caused the doors to open during the collision, resulting in the occupants’ being ejected from the van, and by other defects in the automobile.5
After a four-week trial, the late Judge Dooling submitted the case to the jury upon twenty-five interrogatories. The jury found, inter alia: (1) that Valentine’s negligence was the proximate cause of the accident; (2) that the door handle design was defective; (3) that the Caiazzos were ejected from the van because of the defect; (4) that the Caiazzos’ injuries were aggravated beyond the injuries they would have sustained if there had been no defect; (5) that Turi Caiazzo’s total damages were $750,000, of which $500,000 (or two-thirds of her total damages) represented the aggravation of her injuries due to defective door handle design; (6) that Frank Caiazzo’s total damages were $200,000, of which $150,000 (or three-fourths of his total damages) represented the aggravation of his injuries due to the defective door handle design; and (7) that the Caiazzos’ injuries would have been reduced by 25% if they had been wearing seat belts. In sum, the jury concluded that Turi Caiazzo was entitled to recover $562,-500 against Valentine and $375,000 of that amount against VWAG jointly.6 Frank Caiazzo was entitled to recover $150,000 against Valentine, for $112,500 of which VWAG was found jointly liable.
VWAG moved for judgment notwithstanding the verdict and a new trial and renewed its motion for a directed verdict. Judge Dooling denied these motions, but proposed a remittitur of $100,000 with respect to the $200,000 awarded to Frank Caiazzo and $350,000 with respect to the $750,000 awarded to Turi Caiazzo, before the 25% reduction for the Caiazzos’ failure to wear seat belts. Caiazzo v. Volkswagen-werk, A.G., 468 F.Supp. 593 (E.D.N.Y.1979). This resulted in a final judgment for Frank Caiazzo of $75,000 against Valentine, of which $56,250 was against VWAG jointly, and a judgment for Turi Caiazzo of $300,-0,00 against Valentine, of which $200,000 was against VWAG jointly. The Caiazzos accepted the remittitur and judgment was entered on May 5, 1979.
II. The Issues on Appeal
VWAG raises a variety of issues7 in its appeal: (1) whether there was sufficient [245]*245evidence to support a finding of enhanced injuries or, in the alternative, whether the evidence necessitated a finding that all of plaintiffs-appellees’ injuries would have been mitigated by wearing seat belts; (2) whether the trial court erred in concluding that plaintiffs-appellees did not have the burden of segregating the enhanced injuries from those resulting from the collision; and (3) whether the trial court erred in concluding that plaintiffs-appellees’ failure to wear seat belts could not constitute contributory negligence.8
III. The Legal Background
The second collision doctrine of liability is relatively simple. Since manufacturers are already under a duty to construct vehicles that are free of latent defects, MacPherson v. Buick Motor Co., 217 N.Y. 382, 111 N.E. 1050 (1916), it follows that the manufacturer’s liability for injuries proximately caused by these defects should not be limited to collisions in which the defect causes the accident, but should extend to situations in which the defect caused injuries over and above that which would have occurred from the accident, but for the defective design. Larsen v. General Motors Corp., 391 F.2d 495 (8th Cir. 1968); Bohn v. Triumph Corp., supra.
What appears to be simple in theory, however, presents a myriad of problems in the litigation of this principle. See Foland, Enhanced Injury: Problems of Proof in “Second Collision” and “Crashworthy” Cases, 16 Washburn L.J. 600 (1976). This appeal raises some of the thornier issues: (1) the extent of the plaintiff’s burden of proof regarding causation, i. e., to what degree must the plaintiff segregate the enhancement injuries from the primary injuries; (2) the effect of the plaintiff’s contributory fault, in particular, the plaintiff’s failure to use a seat belt, on the plaintiff’s claim that the automobile is defective, where the defendant argues that the “second collision” would not have occurred had seat belts been used; and (3) the interrelation between (1) and (2).
The problem with the litigation of crash-worthiness cases generally is that, except for the instance in which the injuries caused by the design defect are clearly distinguishable from those caused by the initial collision (e. g., burn injuries from the explosion of a defectively designed gas tank, as in Turcotte v. Ford Motor Co., 494 F.2d 173 (1st Cir. 1974)), the evidence will necessarily include a determination of what might have happened in the collision under different circumstances. The problem, as the trial court pointed out, is that “[w]hat might have happened is rarely susceptible of proof.” Caiazzo v. Volkswagenwerk, A.G., supra, 468 F.Supp. at 598.
In this case, the design defect and seat belt issues add considerably to the jury’s tasks. Not only must the jury attempt to determine what injuries each plaintiff [246]*246would have suffered if the doors had not opened, but it must also determine what enhanced injuries each suffered by reason of the doors opening and what part of the enhanced injuries, if any, would have been avoided had they been wearing seat belts.9
IV. The Trial Court’s Approach
The trial court ruled that the Caiazzos need prove only what was referred to as the “fact” of enhancement — “that the defect in design was a factor in causing [some of] plaintiffs’ injury and damage.” Caiazzo v. Volkswagenwerk, A.G., supra, 468 F.Supp. at 601. VWAG, in seeking to limit its liability to the enhanced damages, had the burden of apportionment, i. e., it had to segregate the injuries between those caused by the collision, by the alleged design defects,10 and by the nonuse of the seat belts.
While we appreciate the difficulties of predicting how the New York courts would deal with' these intertwining issues, we think the trial court’s approach in unraveling these problems places too heavy a burden on the manufacturer and contradicts the theoretical underpinnings of the second collision doctrine. Placing on VWAG the burden of apportioning between the collision injuries and the enhanced injuries, in addition to the burden of proving what injuries would have been reduced if the Caiazzos had worn their seat belts, corners the manufacturer into offering evidence of a plethora of hypothetical and speculative possibilities. For purposes of apportioning damages, VWAG would have had to distinguish between the types of injuries the Caiazzos would have incurred if: (1) they were wearing seat belts and the doors did not open; (2) they were not wearing seat belts and the doors did not open; (3) they were wearing seat belts and the doors did open; and (4) they were not wearing seat belts and the doors did open (which appears to be what occurred in this case).11
The upshot is that the manufacturer would be forced to prove a part of plaintiffs-appellees’ ease. A simpler, fairer, and conceptually more sound approach, is to require the Caiazzos to establish the extent of enhanced injuries attributable to the defective design.
With respect to the seat belt issue, the burden of proving the consequences of the Caiazzos’ failure to wear seat belts was properly placed with VWAG. We affirm this part of the district court’s decision as well as its instruction to the jury that the seat belt issue could be considered only in assessing damages. See Spier v. Barker, 35 N.Y.2d 444, 323 N.E.2d 164, 363 N.Y.S.2d 916 (1974).
V. The Second Collision Doctrine
The seminal case for the second collision doctrine is Larsen v. General Motors Corp., [247]*247supra. Larsen involved a head-on collision in which the steering mechanism in the plaintiff’s vehicle displaced rearward and struck the plaintiff in the head. The plaintiff alleged that, while the defective steering assembly did not cause the accident, it exacerbated the injuries he would have suffered in the collision absent the defective assembly. General Motors, the manufacturer of the vehicle, argued that it had “no duty” to design an automobile that is “ ‘safe’ or ‘safer’ to occupy during collision impacts.” 391 F.2d at 497. The district court, relying on Evans v. General Motors Corp., 359 F.2d 822 (7th Cir.), cert. denied, 385 U.S. 836, 87 S.Ct. 83, 17 L.Ed.2d 70 (1966), granted summary judgment for General Motors. In Evans, the Seventh Circuit ruled that manufacturers had no duty to make automobiles crashworthy, id. at 824, because collisions are not an intended use of automobiles. In Larsen, the Eighth Circuit, rejecting the Evans approach, ruled that, since collisions are a foreseeable, though unintended, result of the normal operation of automobiles, a manufacturer “is under a duty to use reasonable care in the design of its vehicle to avoid subjecting the user to an unreasonable risk of injury in the event of a collision.” 391 F.2d at 502. Larsen did not hold, however, that manufacturers are under a duty to design “accident-proof or foolproof” automobiles, id. at 502, nor did it hold that manufacturers are “insurers.” Id. at 503. In the flurry of litigation that followed the decision, some jurisdictions accepted or expanded Larsen, while others adopted the Evans approach.12 Foland, supra, at 600-01.
The New York courts addressed the issue in Bolm v. Triumph Corp., supra. Bolm involved a collision between an automobile and a motorcycle. Plaintiff, the motorcyclist, in addition to suing the owner and operator of the automobile, sued the manufacturer of the motorcycle, alleging that the placement of the metal luggage rack on top of the motorcycle’s gas tank in front of the saddle constituted a design defect that exacerbated plaintiff’s injuries. The New York Court of Appeals, rejecting a prior New York decision that held that second collision injuries were not actionable, Edgar v. Nachman, 37 A.D.2d 86, 323 N.Y.S.2d 53 (3d Dep’t 1971), adopted the second collision theory of liability, ruling that manufacturers’ liability extended to “unreasonably dangerous (latent) design defects which enhance or aggravate injuries.” Bolm, supra, 33 N.Y.2d at 158, 305 N.E.2d at 773, 350 N.Y.S.2d at 650.13
Neither Bolm nor subsequent New York decisions provide guidance, however, as to exactly how these cases are to be litigated. Specifically, they do not answer the ques[248]*248tions presented in this case: What causal link must a plaintiff establish between the defect and the injuries? What bearing will a plaintiffs own failure to use available safety devices have on his claim that the vehicle is unsafe? Before we consider these questions, however, we must first address VWAG’s claim regarding sufficiency of the evidence.
VI. Sufficiency of the Evidence
VWAG contends that, regardless of the resolution of these various burden of proof issues, it is entitled to a directed verdict either because the Caiazzos have failed to offer any evidence of enhancement, or, in the alternative, because any injuries attributable to the defect were barred by the seat belt defense.
With respect to the verdict, the trial court commented that “[a]n allocation of the kind which the jury made ... does not reveal its logic, and cannot be pinned to specific items of evidence.” Caiazzo v. Volkswagenwerk, A.G., supra, 468 F.Supp. at 598. We agree. We also agree, dubi-tan te, with the trial court’s conclusion that the Caiazzos submitted sufficient evidence of enhancement, which would have occurred even if they were wearing seat belts, to create a jury issue. We also concur with the court’s decision that the evidence on nonuse of seat belts does not necessarily mandate a finding of total mitigation of the Caiazzos’ enhanced injuries.
The witnesses on the design defect and causation issues for the Caiazzos were William Steiglitz, an accident reconstruction expert, and Dr. Norman Silver, the Caiaz-zos’ treating physician. In addition to his testimony regarding the manner of ejection and the kinematic effects of the accident sequence, Steiglitz, in response to hypothetical questions, testified that the defects in the doors were substantial factors in causing the injuries sustained by the Caiaz-zos. The weight to be accorded to his testimony was within the province of the jury to decide. See e. g., Manning v. New York Telephone Co., 388 F.2d 910, 912 (2d Cir. 1968).
Dr. Silver, in response to two hypothetical questions based on the testimony of Frank Caiazzo, testified that some of his injuries occurred between the time that Frank Caiazzo said he saw “the ground coming up at him” and the end of the accident sequence. Thus, the trial court noted:
The jury could have accepted or rejected an expert’s opinion that Frank Caiazzo’s compression fracture was more likely to have occurred when he was thrown from the van rather than when he was in it. The jury could have accepted or rejected an expert opinion that the fracture of plaintiff Turi Caiazzo’s right shoulder blade came about from contact with the seat back and the bulkhead behind it, and could, with more confidence, have found that her right foot injuries were brought about from some kind of twisting contact with the pedals on the floor of the van in front of the driver’s seat. The jury had before it an opinion that the hematoma over the occipital region of Turi Caiazzo’s head (at the back of the skull) could have resulted either from her head striking the door frame or the ground; the same expert indicated that the left ankle injury could have occurred when plaintiff Turi Caiazzo was thrown from the van.
Record at A206-07.
Turning to the effect of the Caiazzos’ failure to wear seat belts, it is not disputed that they would not have been ejected had they been wearing them. There was some evidence, however, to the effect that the Caiazzos might have incurred some of their enhanced injuries, or possibly other injuries even if they had been belted during the course of the roll-over. VWAG did not offer uncontradicted evidence to establish that no reasonable jury could find that the Caiazzos’ use of the lap belts would have totally mitigated the enhanced injuries attributable to the opening of the doors. The jury found that both the initial injuries and the enhanced injuries would have been reduced by the same proportionate amount (25%) if seat belts had been worn, but there was no evidence of (and the jury was not [249]*249asked to determine) what enhanced injuries, resulting from the opening of the doors, would have occurred within the car had the Caiazzos been wearing belts.
We are not persuaded, however, that the jury’s apportionment of damages between the initial accident, the design defect, and the failure to wear seat belts, is consistent with the evidence. The most serious consequence of the design defect was that the Caiazzos were ejected from the van. Both the Caiazzos’ expert and VWAG’s expert, however, agreed that the Caiazzos would not have been ejected had they been wearing their seat belts.14 The jury nevertheless concluded that wearing the seat belts would have reduced the injuries caused by the design defect by only 25% even though the most serious consequence of the design defect, the ejection from the van, would not have occurred. This conclusion simply does not make sense. Another problem with the verdict is the disproportionately high percentage of Turi Caiazzo’s injuries attributed to the design defect. The evidence indicates that the most serious of her injuries, the ankle fractures, occurred inside the van. The jury concluded, however, that two-thirds of her injuries were caused by her being ejected from the van. But there was no evidence that her ankle fractures, her most serious and painful injuries, resulted from the ejection. See note 9 supra.
We need not rule on the trial court’s valiant attempt to rationalize these inconsistencies in the jury verdict (which it acknowledged was “inscrutable”), since we are granting a new trial on other grounds. Since we have concluded that there was sufficient evidence to submit the case to the jury, we now must consider how the jury should have been instructed as to the burden of proof in a second collision situation and as to the legal effect of the failure to use seat belts.
VII. The Burden of Proof
Decisions following Larsen have treated the plaintiff’s burden of proof in two different ways. One group of decisions, the most notable of which is Huddell v. Levin, 537 F.2d 726 (3d Cir. 1976), requires a plaintiff to distinguish between those injuries that are attributable to the accident and those that are attributable to the alleged design defect. See Stonehoeker v. General Motors Corp., 587 F.2d 151, 158 (4th Cir. 1978) (“the burden is on the plaintiff either to establish that he would have suffered no injury or the extent of the injury he would have suffered, had the vehicle been properly designed”); Jeng v. Witters, 452 F.Supp. 1349, 1361 (M.D.Pa. 1978), aff’d without opinion, 591 F.2d 1335 (3d Cir. 1979); Yetter v. Rajeski, 364 F.Supp. 105, 109 (D.N.J.1973); Foland, supra.
Other jurisdictions have ruled that a plaintiff need not prove the nature and extent of the enhanced injuries, Fox v. Ford Motor Co., 575 F.2d 774 (10th Cir. 1978),15 but need only offer “[s]ome evidence of enhancement” to present a jury issue. Lahocki v. Contee Sand & Gravel Co., 41 Md.App. 579, 398 A.2d 490, 501 (Ct. Spec.App.1979), rev’d on other grounds sub nom. General Motors Corp. v. Lahocki, 286 Md. 714, 410 A.2d 1039 (1980) (emphasis in original). See Huddell v. Levin, supra, 537 F.2d at 744-47 (concurring opinion). See generally Note, Apportionment of Damages in the “Second Collision” Case, 63 Va.L.Rev. 475 (1977); Comment,- Automobile Design [250]*250Liability: Larsen v. General Motors and Its Aftermath, 118 U.Pa.L.Rev. 299 (1969).16
Judge Dooling, relying on a statement in Bolm v. Triumph Corp., 83 N.Y.2d 151, 305 N.E.2d 769, 350 N.Y.S.2d 644 (1973), that the manufacturer’s liability is to be determined by the traditional rules of negligence and strict liability, applied the Restatement (Second) of Torts §§ 433A and 433B (1965) and concluded that VWAG had the burden of apportionment of the injuries. Caiazzo v. Volkswagenwerk, A.G., supra, 468 F.Supp. at 602-04. That language in Bolm, however, refers to determining the threshold question of whether a manufacturer is liable at all; it does not concern what damages the manufacturer will be liable for, nor does it frame the elements of a plaintiff’s prima facie case.
We do not agree that it is adequate to prove only the “fact” of damage and, instead, adopt, in principle, the approach taken by the Third Circuit in Huddell v. Levin, supra. Huddell involved an allegedly defective head restraint. Plaintiff’s decedent’s head struck the head restraint after his car was hit in the rear. Plaintiff claimed that the head restraint had an unusually sharp edge, which caused plaintiff’s decedent’s death when it struck him on the rear of the head. At trial, plaintiff offered proof that the head restraint was defective and that the defect was a proximate cause of death. The Third Circuit ruled that this proof was insufficient to make out a prima facie claim under a second collision theory of liability and held:
[Cjrashworthy or second collision cases impugning the design of an automobile require a highly refined and almost invariably difficult presentation of proof in three aspects. First, in establishing that the design in question was defective, the plaintiff must offer proof of an alternative safer design, practicable under the circumstances .... Second, the plaintiff must offer proof of what injuries, if any, would have resulted had the alternative, safer design been used.... Third, ... the plaintiff must offer some method of establishing the extent of enhanced injuries attributable to the defective design.
537 F.2d at 737-38.
While we do not predict that New York would adopt this mechanical formula of proof, the plaintiff should be required to prove the extent of the enhanced injuries attributable to the defective design, particularly in a case such as this, where the wearing of seat belts would have eliminated most, if not all, of the enhanced injuries resulting from the design defect. This evidence will generally, perhaps even necessarily, be in the form of expert testimony. See Yetter v. Rajeski, supra, 364 F.Supp. at 109; Horn v. General Motors Corp., 17 Cal.3d 359, 551 P.2d 398, 131 Cal.Rptr. 78 (1976); Foland, supra, at 616-19.
To require proof merely of the fact of enhancement permits a jury to engage in undue speculation as to the causes of various injuries and gives a jury dangerous latitude in assigning responsibility to the defendant who appears most able to pay a [251]*251plaintiff’s award. The case at bar illustrates the problem. While we have concluded that there was probably sufficient evidence of the fact of enhancement to withstand a motion for directed verdict, the evidence on this point was quite thin. The Caiazzos presented no evidence as to what injuries would have resulted from the impact had there been no design defect. Moreover, as the trial court noted, the expert testimony could “easily” have done more “to elucidate the mechanics of the accident, to indicate the different reconstructions of the sequence that might be inferred, and to describe the physical injuries sustained by each plaintiff and the damaging forces that could be inferred from those injuries.” Caiazzo v. Volks-wagenwerk, A.G., supra, 468 F.Supp. at 598. The expert evidence was to the effect that various of the Caiazzos’ injuries could have been sustained either from the alleged design defects or from the primary collision. The Caiazzos’ expert suggested that even belted riders could sustain crushed skulls in a roll-over, but he did not establish whether belting could have prevented distinctly identifiable injuries resulting from the opening of the doors. Without evidence as to what injuries would have resulted had the Caiazzos not been ejected, the jury had to speculate and hypothesize as to which injuries resulted from the defective door design. The disharmony between the jury’s verdict and the evidence, see discussion supra, exemplifies the problem with this approach.
We realize that a plaintiff’s burden of offering evidence of what injuries would have resulted absent the alleged defect will be heavy in some instances and perhaps impossible in others. Where it is impossible, however, the plaintiff has merely failed to establish his prima facie case, i. e., that it is more probable than not that the alleged defect aggravated or enhanced the injuries resulting from the initial collision. Moreover, in those instances in which the plaintiff cannot offer any evidence as to what would have occurred but for the alleged defect, the plaintiff has not established the fact of enhancement at all.
VIII. Legal Effect of the Failure to Wear Seat Belts
The starting point of any discussion of the seat belt defense under New York law is Spier v. Barker, supra. Spier was a negligence action brought when (as at the time of the accident in the instant action) New York was a contributory negligence jurisdiction. The New York Court of Appeals ruled that a plaintiff’s nonuse of seat belts may be considered only in determining damages and that failure to use a seat belt constitutes neither contributory negligence per se nor contributory negligence as a matter of common law. 35 N.Y.2d at 450-51, 323 N.E.2d at 167-68, 363 N.Y.S.2d at 921. The court noted that “the doctrine of contributory negligence is applicable only if the plaintiff’s failure to exercise due care causes, in whole or in part, the accident, rather than when it merely exacerbates or enhances the severity of his injuries,” id. at 451, 323 N.E.2d at 168, 363 N.Y.S.2d at 921 (emphasis in original), and added the caveat that “[n]ot involved in this case, and not considered, is an issue in which the failure to wear a seat belt is an alleged cause of the accident.” Id. n.3.
VWAG urges us to carve out an exception to the Spier “mitigation rule” for second collision cases. It argues that (1) Spier does not apply to strict products liability claims; and (2) this is the exceptional case in which nonuse of the seat belt causes the accident. We disagree with both of these positions.
The doctrine of strict products liability in New York dates from Codling v. Paglia, 32 N.Y.2d 330, 298 N.E.2d 622, 345 N.Y.S.2d 461 (1973). See Bolm v. Triumph Corp., supra, 33 N.Y.2d at 159-60, 305 N.E.2d at 773, 350 N.Y.S.2d at 650. Codling provides:
[Ujnder a doctrine of strict products liability, the manufacturer of a defective product is liable to any person injured or damaged if the defect was a substantial factor in bringing about his injury or damages; provided: (1) that at the time of the occurrence the product is being [252]*252used (whether by the person injured or damaged or by a third person) for the purpose and in the manner normally intended, (2) that if the person injured or damaged is himself the user of the product he would not by the exercise of reasonable care have both discovered the defect and perceived its danger, and (3) that by the exercise of reasonable care the person injured or damaged would not otherwise have averted his injury or damages.
32 N.Y.2d at 342, 298 N.E.2d at 628-29, 345 N.Y.S.2d at 469-70.
VWAG maintains that the Caiazzos’ failure to use their seat belts constitutes a failure to exercise reasonable care, which, in turn, bars them from recovery under part (3) of the Codling test. New York courts have uniformly interpreted the “exercise of reasonable care” language to be identical to that used for the doctrine of contributory negligence. See, e. g., Bolm v. Triumph Corp., 71 A.D.2d 429, 434, 422 N.Y. S.2d 969, 973 (4th Dep’t 1979). There is nothing in Spier v. Barker, supra, decided after both Codling and Bolm, that indicates that the Spier mitigation rule would not be applicable to a second collision products liability case.
Moreover, there is nothing in Spier to support VWAG’s alternative argument that this is a case in which the failure to wear seat belts causes the accident. See Spier, supra, 35 N.Y.2d at 451, 323 N.E.2d at 168, 363 N.Y.S.2d at 921. VWAG argues that, as to them, “the accident” is the Caiazzos’ ejection from the van. The “accident” quite simply is the collision between the two vehicles. VWAG, of course, is liable only for the injuries resulting from the defective door latch, which were certainly enhanced by the ejection. As Judge Dool-ing correctly instructed the jury, to the extent that VWAG establishes that the enhancement would not have occurred if plaintiffs had worn seat belts, the extent of any recovery against it must be diminished accordingly.
IX. Issues on Retrial
For the reasons already set forth, the case must be remanded for a new trial. The use of the special interrogatories at the first trial, however, obviates the need for an entire new trial and permits the second trial to be limited to the damages issue.
Specifically, the jury determined at the first trial that only the door latch assembly was defectively designed and that the Caiazzos were ejected and were not wearing seat belts. These issues need not be retried.
The Caiazzos will be required to establish the extent of enhanced injuries attributable to the defective design. This issue will be less complicated at the second trial, since the factual dispute as to what actually occurred during the accident has already been resolved.17 VWAG will have the burden of establishing what injuries would have been obviated had the Caiazzos been wearing their seat belts.
Special interrogatories should be put to the jury to determine whether the Caiazzos would have incurred some injuries from the opening of the doors even if plaintiffs-ap-pellees had been wearing seat belts. If the jury answers in the affirmative, the Caiaz-zos have established enhancement. The jury should then be instructed to'apportion the Caiazzos’ injuries among those caused by the collision, those caused by the defective door latch assembly, and those caused by the nonuse of seat belts. The extent to which liability should be reduced by the Caiazzos’ negligence (in not wearing seat belts) should be appraised separately for Valentine and for VWAG.18
Affirmed in part, reversed in part, and remanded for proceedings consistent with this opinion.