This diversity action is a personal injury suit arising from an automobile accident that occurred in Milwaukee, Wisconsin. Plaintiffs tried the case to a jury on the theories of negligence and products liability. At the close of the evidence, defendants moved for a directed verdict on both claims. The district court allowed the products liability claim to. go to the jury, but directed a verdict for defendants on the negligence claim. The jury found for defendants on the products liability claim. Plaintiffs moved for a new trial, challenging the directed verdict on the negligence claim as erroneous. The motion was denied, and the plaintiffs appeal. We reverse the judgment of the district court.
*640I.
The automobile accident that gave rise to this suit occurred at approximately 2:00 a.m. on October 30, 1981, at the intersection of 24th and Wells Streets in Milwaukee, Wisconsin. Plaintiff Lauren Chaulk was the front seat passenger in a 1977 Volkswagen Rabbit. As the Volkswagen entered the intersection, it was broadsided on the passenger side by a 1978 Ford LTD. The drivers of both cars sustained minor injuries. Chaulk, however, was ejected from the car and thrown to the ground some 65 feet away from the estimated point of impact. She was knocked unconscious and remained in a coma for several months.
At trial, plaintiffs argued that the door latch system of the 1977 Volkswagen Rabbit was designed in such a way that the force of a side impact collision could release the latch, and that Chaulk was ejected from the car as a result of just such an inadvertent release. In plaintiffs’ view, this latch system was both negligently designed and inherently dangerous.
At the close of the evidence, defendants moved for a directed verdict on both the negligence and the products liability claims. The court granted the motion as to the negligence claim, but permitted the products liability claim to go to the jury. On September 26, 1985, the jury returned a verdict in favor of defendants on the products liability claim. Plaintiffs filed a motion for a new trial on October 4, 1985. The district court denied this motion on December 13, 1985, and the plaintiffs have appealed.
II.
A.
In diversity cases, this court applies the state standard of review to the trial court’s decision to grant or deny a directed verdict. Gonzalez v. Volvo of America Corp., 752 F.2d 295, 301 (7th Cir.1985); Davis v. FMC Corp., 771 F.2d 224, 229 (7th Cir.1985); McMahon v. Eli Lilly & Co., 774 F.2d 830, 832 (7th Cir.1985). Because the law of Wisconsin governs in this case, we must apply the Wisconsin standard. In Wisconsin, the standard of review for an appellate court when passing on the correctness of the trial court’s decision to direct a verdict is “whether the trial court was clearly wrong.” State v. Leach, 124 Wis.2d 648, 370 N.W.2d 240, 249 (1985); Greiten v. LaDow, 70 Wis.2d 589, 235 N.W.2d 677, 683 (1975).
A Wisconsin trial court should not direct a verdict if there is “any credible evidence which under a reasonable view would support a verdict contrary to that which is sought.” State v. Leach, 370 N.W.2d at 249 (quoting Thompson v. Howe, 77 Wis.2d 441, 253 N.W.2d 59, 62 (1977)).1 In this case, plaintiffs’ evidence of negligence was not only credible but substantially uncontroverted by defendants. We therefore conclude that the trial court’s decision to direct the verdict on the negligence claim in favor of defendants was “clearly wrong.”2
*641B.
In design defect cases, “[a]ll that it is necessary to prove [negligence] is that the product is designed with a lack of ordinary care and that lack of care resulted in injury.” Greiten, 235 N.W.2d at 685; see also Fischer v. Cleveland Punch & Shear Works, 91 Wis.2d 85, 280 N.W.2d 280, 283 (1979) (“A cause of action in negligence requires proof that the defendant failed to exercise ordinary care and that the act or omission complained of was the cause of the plaintiff’s injury.”). “Ordinary care" involves the concept of foreseeability. Greiten, 235 N.W.2d at 685; Fischer, 280 N.W.2d at 283. Therefore, for plaintiffs to be entitled to a new trial on the issue of negligence, the record must clearly show credible evidence as to (1) a lack of ordinary care, including the foreseeability of the accident; (2) causation; and (3) injury.3 We hold that the plaintiffs presented some credible evidence as to each of these issues.4
*642That there was some credible evidence as to the issues of causation and injury is not seriously disputed here. At trial, plaintiffs’ negligence case rested on the testimony of one expert witness, Mr. Jack Martens. Martens, an engineer with expertise in automobile safety, testified that in his opinion the cause of Chaulk’s ejection from the car, and therefore the cause of her injuries, was a negligently designed latch system. The latch system is designed so that in theory, if the door were pushed inward in a side impact collision, the inward movement could release the latch. In Martens’ opinion, such an inadvertent release caused Chaulk’s ejection from the automobile. Although defendants presented opposing testimony that Chaulk was ejected through the passenger window rather than the door, and that the door only opened after her exit, on appeal they do not argue that plaintiffs’ evidence as to causation was incredible. Chaulk’s injuries are similarly not in dispute.
The central issue in this case is therefore whether Martens presented credible evidence as to Volkswagen’s failure to exercise ordinary care. We find that Martens’ evidence was not only credible but uncontroverted by the defendants. Martens testified that in his opinion Volkswagen was negligent in designing the 1977 Rabbit latch system. He stated that a reasonable design engineer should have foreseen both the possibility of a side impact collision, and the possibility of inadvertent release when this latch system was struck in such a collision. He testified that alternate, safer designs were available at the time of production of this particular Volkswagen model. Martens further stated that later in the 1977 model year — after the model in question was designed — Volkswagen changed its door handle and door latch design. In Martens’ opinion, this new latch would not have activated in the same kind of collision.
The defendants argue, nevertheless, that plaintiffs presented insufficient evidence of a failure to exercise ordinary care, and point to Martens’ testimony under cross-examination. Under cross-examination, Martens conceded that even the alternate designs he presented at trial were defective to some degree. Upon further questioning, Martens took the position that any means that can inadvertently open a latch produces a defect in the latch. He also conceded that any latch will fail if enough force is put on it. Therefore, the defendants argue that the impact of Martens’ testimony was only that some latch systems are safer than others, but all are unsafe to some degree. In defendants’ view, this evidence fails to establish a departure from ordinary care on Volkswagen’s part.
If Martens’ testimony amounted to nothing more than this, the defendants might be right. Contra Sumnicht v. Toyota Motor Sales, U.S.A., Inc., 121 Wis.2d 338, 360 N.W.2d 2, 17 (1984) (an entire industry may be found negligent); D.L. By Friederichs v. Huebner, 110 Wis.2d 581, 329 N.W.2d 890, 907 (1983) (same). In fact, however, Martens’ testimony shows that he distinguished at least two kinds of latch system defects. One type of defect becomes apparent in a side impact, where the point of collision is the rear width of the car — that is, the length of the car between the door and the rear end — rather than the door itself. In such a situation, the body of the car may stretch, pulling the latch away from the sheet metal to which it is mounted and forcing the release mechanism open. As to this type of latch system defect, Martens indeed faulted almost every car he was asked to consider, including the alternate latch designs he presented as safer than the design in the 1977 Volkswagen Rabbit.
However, in the accident that led to Chaulk’s injury the point of impact was the door itself. The latch was activated by direct force and not by stretching. Although Martens again faulted many cars with respect to this second type of design defect, he also clearly testified that other, safer designs were available at the time that would not have given way in a side impact collision, and that a reasonable design engineer should have foreseen such an *643accident. This testimony was not substantially shaken during cross-examination, nor was plaintiffs' evidence controverted or refuted by opposing testimony that this type of collision was not foreseeable or that alternate designs would not have been safer.5
On facts such as these — where the plaintiffs’ evidence as to each element of a negligence claim under Wisconsin law was not only “credible” but substantially unopposed — we must conclude that the trial court was “clearly wrong” in directing a verdict for defendants. The plaintiffs were entitled to submit their negligence claim to the jury. We, of course, express no opinion as to the merit of the claim.
The decision of the trial court is reversed and the cause remanded for a new trial on the issue of negligence.
. The Leach court further stated:
A case should be taken from the jury and a verdict directed against a party only when the evidence gives rise to no dispute as to the material issues or only when the evidence is so clear and convincing as reasonably to permit unbiased and impartial minds to come to but one conclusion____ [The] court must take that view of the evidence which is most favorable to the party ... against whom the verdict was sought to be directed____ If there is any evidence other than mere conjecture or incredible evidence to support a contrary verdict, the case must go to the jury.
State v. Leach, 370 N.W.2d at 249 (quoting Zillmer v. Miglautsch, 35 Wis.2d 691, 151 N.W.2d 741, 745 (1967)) (citations omitted). See also Anderson v. Joint School Dist., 24 Wis.2d 580, 129 N.W.2d 545, 547 (1964).
. The parties argue at length about the implications for this case of Hansen v. Cessna Aircraft Co., 578 F.2d 679 (7th Cir.1978). In Hansen we held that, under Wisconsin law, when a plaintiff has made out a prima facie case for both a negligence and a products liability claim, she cannot be made to elect between them. Rather, both claims must be submitted to the jury. Hansen, 578 F.2d at 685.
As we noted in Hansen, the practice of submitting a case on both negligence and products liability theories had caused confusion because of the unique development of Wisconsin tort law. When the Wisconsin Supreme Court *641adopted section 402A of the Restatement (Second) of Torts, it termed the new theory of recovery “negligence per se” in order to make clear that any recovery under products liability would be subject to Wisconsin’s comparative negligence statute. Because there were then apparently two kinds of "negligence" theories — ordinary and “per se” — some Wisconsin courts began to require plaintiffs to choose between them. See Hansen, 578 F.2d at 682-83. The Wisconsin Supreme Court disapproved this practice in Howes v. Deere & Co., 71 Wis.2d 268, 238 N.W.2d 76 (1976), and distinguished the two types of theories. 238 N.W.2d at 79; Hansen, 578 F.2d at 683. See also D.L. By Friederichs v. Huebner, 110 Wis.2d 581, 329 N.W.2d 890, 903 (1983) ("plaintiffs need not elect between presenting the two theories to the jury”); Fischer v. Cleveland Punch & Shear Works Co., 91 Wis.2d 85, 280 N.W.2d 280, 286 (1979); Greiten v. LaDow, 70 Wis.2d 589, 235 N.W.2d 677, 685 (1975). Hansen made clear that, in light of these cases, submission of both claims was required by Wisconsin law, and not a matter of trial court discretion. All Hansen adds to the present case is additional authority for the proposition that the trial judge here was required to submit the negligence claim to the jury in addition to the products liability claim, if plaintiffs had presented some credible evidence as to each element of the negligence claim. Thus, our inquiry remains the same: whether the trial court was clearly wrong in concluding that plaintiffs had presented no credible evidence to support a negligence claim. If the plaintiffs did present such credible evidence, the trial judge was required to submit their claim to a jury.
. Plaintiffs characterize the essential elements of a design defect case as (1) a dangerous design; (2) knowledge on the part of the manufacturer; and (3) a practical alternative. But the Wisconsin courts have not had occasion to pass on this formulation, and we therefore express no opinion as to whether this characterization is correct.
. It might be argued that the jury's conclusion that Volkswagen was not strictly liable — that is, that there was no defect in the 1977 Volkswagen Rabbit latch system that was unreasonably dangerous — precludes a finding that Volkswagen was negligent. In Wisconsin, strict liability was adopted in order to "aid the plaintiff in proving the case.” D.L. By Friederichs v. Huebner, 110 Wis.2d 581, 329 N.W.2d 890, 903 (1983); see also Burrows v. Follett & Leach, Inc., 115 Wis.2d 272, 340 N.W.2d 485, 491 (1983) ("One of the principal effects of the rule of strict liability is that it relieves the plaintiff of the burden of proving specific acts of negligence.”); Howes v. Deere & Co., 71 Wis.2d 268, 238 N.W.2d 76, 79 (1976) (strict liability relieves plaintiff of proving specific acts of negligence); Greiten v. LaDow, 235 N.W.2d at 684 (“The intent of [the products liability] rationale is to permit a plaintiff to recover under such circumstances where it is impossible or unduly burdensome to show the elements of negligence.’’). We might therefore conclude that because a products liability claim is easier to prove than a negligence claim, if a products liability claim fails a negligence claim arising out of the same conduct must necessarily fail. If this reasoning were correct, the trial court’s directed verdict on the negligence count would be at worst harmless error.
However, the Wisconsin courts have repeatedly rejected this reasoning, and held that a verdict finding the defendant negligent but not strictly liable is not inconsistent. See Giese v. Montgomery Ward, Inc., 111 Wis.2d 392, 331 N.W.2d 585, 596 (1983); Fischer v. Cleveland Punch & Shear Works, 280 N.W.2d at 286; Howes v. Deere & Co., 238 N.W.2d at 79-80; Greiten v. LaDow, 235 N.W.2d at 685. But see Flaminio v. Honda Motor Co., 733 F.2d 463, 466-67 (7th Cir.1984) (in diversity case under Wisconsin law, difference between the elements of strict liability and negligence was so small that failure to give explicit strict liability instruction was not reversible error in design defect case). Under Wisconsin law, in this case it was possible for plaintiff to recover for the negligent design of the 1977 Volkswagen Rabbit passenger door latch system, even though that system was not found to be defective and unreasonably dangerous for the purpose of a products liability claim.
. Defendants did present evidence that there had been no previous litigation involving inadvertent release of the 1977 Volkswagen Rabbit latch. But this alone cannot be sufficient to negate evidence of negligence.