Appellants, John and Carol Baker, filed a complaint against appellee, T.V. Morrison, alleging appellee attempted to pass their vehicle on a highway and negligently struck their vehicle from behind while they were making a left turn. The suit was tried to a St. Francis County jury which returned general verdict in favor of appellee. As their sole point for reversal, appellants contend the trial court erred in denying their motion in limine to exclude evidence concerning their failure to wear seat belts at the time of the accident. We agree that the denial of the motion in limine was error and reverse.
On appeal, appellants argue the evidence of their failure to wear seat belts is not relevant to the type of injuries they sustained. They also argue appellee failed to prove that their failure to wear seat belts was a proximate cause of their injuries. Appellee responds with the claim that appellants’ failure to wear seat belts was evidence of their comparative fault and admissible as such.
We note that in appellee’s brief, in addition to his argument that appellants’ nonuse of seat belts is evidence of comparative *459fault, he asserts that in the absence of medical testimony of proximate cause, a reasonable juror could have concluded that appellants’ injuries would have been reduced or prevented by the use of seat belts. This assertion seems to be an argument that appellants’ nonuse of seat belts is admissible evidence of failure to mitigate damages. However, appellee did not argue this mitigation theory to the trial court in defense of the motion. Furthermore, appellee cites no authority nor presents us with any convincing argument to affirm, on the basis of the mitigation theory, the trial court’s denial of the motion.
Recognition of the seat belt defense on the mitigation theory requires the consideration of issues distinct from those considered under the comparative fault theory. For example, the mitigation theory requires the consideration of the doctrine of avoidable consequences and whether the plaintiff should have anticipated the defendant’s negligence before the accident occurred, while the comparative fault theory, applying Arkansas law on proximate cause, does not. As the trial court did not rule on the mitigation theory, nor was the theory briefed to us on appeal, we. will not address the issue of whether appellants’ nonuse of seat belts was admissible as proof of their failure to mitigate damages.
Thus, we are presented with a very limited issue — whether appellants’ failure to wear seat belts is admissible evidence of their comparative fault. We have never addressed the issue of whether the failure to wear an available seat belt constitutes comparative negligence. We note, however, that the legislature has spoken on this precise issue with the passage of 1991 Ark. Acts 562, the Mandatory Seat Belt Use Act. Specifically, Ark. Code Ann. § 27-37-703 (Supp. 1991) provides that the failure to “use a seat belt shall not be considered under any circumstances as evidence of comparative or contributory negligence, nor shall such failure be admissible as evidence in the trial of any civil action with regard to negligence.” As the date of the accident in question was December 17,1988, it occurred prior to the passage of the Mandatory Seat Belt Use Act. The Act is therefore not applicable to this case.
We begin our analysis with a discussion of the seat belt defense and its existence in other jurisdictions. As seat belts became standard equipment on motor vehicles in the late 1960’s, *460defendants in actions for damages resulting from motor vehicle collisions began raising the issue of a plaintiffs failure to wear an available seat belt. The seat belt arguments presented to various courts over the years have changed according to the changes in negligence law. For example, some of the defense arguments presented have requested a complete bar to the plaintiffs recovery based on the doctrine of contributory negligence. Courts presented with this theory have focused on the liability aspect of negligence and concluded that because the plaintiffs failure to wear a seat belt was not a cause of the collision, the nonuse should not completely bar recovery. See, e.g., Fischer v. Moore, 183 Colo. 392, 517 P.2d 458 (1973). Other arguments presented have requested a reduction in the amount of damages awarded rather than a complete bar to recovery. One such argument is that the evidence of the plaintiffs nonuse of a seat belt should be admissible on the issue of the plaintiffs failure to mitigate his or her damages. Courts presented with the mitigation theory have split. Some have held that the evidence of nonuse is not admissible for various reasons. Britton v. Doehring, 286 Ala. 498, 242 So.2d 666 (1979); Romankewiz v. Black, 16 Mich. App. 119, 167 N.W.2d 606 (1969); Stallcup v. Taylor, 62 Tenn. App. 407, 463 S.W.2d 416 (1970); Carnation Co. v. Wong, 516 S.W.2d 116 (Tex. 1974). Others have held that the evidence of nonuse is admissible as long as the defendant can demonstrate a causal connection between the nonuse and the damages incurred in the collision. Mount v. McClellan, 91 Ill. App. 2d 1, 234 N.E.2d 329 (1968); Spier v. Barker, 35 N.Y.2d 444, 323 N.E.2d 164, 363 N.Y.S.2d 916 (1974). The other theory advanced by defendants requesting a reduction in the amount of damages awarded is that the nonuse should be admissible as evidence of the plaintiffs comparative negligence. Courts presented with this issue have also split. Some courts have held the evidence of nonuse is not admissible because it could not have contributed to the cause of the accident, therefore it is not a valid ground for the defense of comparative fault. Melesko v. Riley, 32 Conn. Supp. 89, 339 A.2d 479 (1975); Amend v. Bell, 89 Wash. 2d 124, 570 P.2d 138 (1977). Other courts have held the nonuse of a seat belt may constitute such comparative negligence as to allow an apportionment of damages; these courts have analyzed the issue in terms of whether there is available evidence that the nonuse contributed to the injuries rather than whether the nonuse contributed to the *461cause of the accident. Law v. Superior Court of Arizona, 157 Ariz. 147, 755 P.2d 1135 (1988); Bentzler v. Braun, 34 Wis. 2d 362, 149 N.W.2d 626 (1967).
Although we have never addressed the issue of the availability of the seat belt defense in a comparative negligence context, we have spoken on the subject in general. In Shelter Mut. Ins. Co. v. Tucker, 295 Ark. 260, 748 S.W.2d 136 (1988), we did not reach the merits of the claim because of the admittedly speculative nature of the physician’s proffered testimony concerning the relation between the injuries sustained and the nonuse of the seat belt. However, we did indicate that evidence concerning a plaintiffs nonuse of a seat belt is relevant to “connect” the nonuse of the belt with the injuries. Id. at 263, 748 S.W.2d at 137. In Harlan v. Curbo, 250 Ark. 610, 466 S.W.2d 459 (1971), we reversed the trial court’s giving of an instruction that when considering the question of the plaintiffs negligence, the jurors could consider the fact that seat belts were available for the plaintiffs’ use. As in the present case, in Harlan, the only evidence of the plaintiffs’ nonuse of seat belts were statements that the, belts were available for use but were not used at the time of the collision. The holding in Harlan was based on the fact that the jury was also instructed on the duty of all persons involved in the accident to use ordinary care for their own safety. The additional reference to the nonuse of seat belts was concluded to have been an unnecessary duplication which was unduly emphasized and may have led the jury to attribute to the plaintiff a greater percentage of total fault than should have been the case. Unlike the present case however, Harlan was a case that was submitted to the jury on interrogatories.
The Eighth Circuit Court of Appeals applied both Tucker, supra, and Harlan, supra, in Benjamin v. Potts, 882 F.2d 1320 (8th Cir. 1989), and predicted that we would allow the admissibility of a plaintiffs failure to wear a seat belt if there is evidence that the failure proximately caused the plaintiffs injuries. Although we think Benjamin correctly applied Arkansas law, we proceed with an analysis of the seat belt defense.
Our comparative fault statute provides that “fault” includes “any act, omission, conduct, risk assumed, breach of warranty, or breach of any legal duty which is a proximate cause *462of any damages sustained by any party.” Ark. Code Ann. § 16-64-122(c) (Supp. 1991). Because this statute defines comparative fault in terms of that which proximately causes damages, we find it necessary to follow the foregoing cases focusing on whether the nonuse caused the plaintiffs injuries, notwithstanding that the nonuse did not cause the accident. Therefore, because appellants’ failure to wear their seat belts was a failure to exercise ordinary care, see Harlan, supra, such failure is not considered “fault” for purposes of comparative fault, unless it was a proximate cause of appellants’ damages. We have defined proximate cause as a cause which, in a natural and continued sequence, produces damage, and without which the damage would not have occurred. Wilson v. Evans, 284 Ark. 101, 679 S.W.2d 205 (1984); AMI 501 (3d ed. 1989). Thus, we hold that appellants’ nonuse of their seat belts may have been admissible as evidence of their comparative fault if such nonuse was a proximate cause of appellants’ injuries.
In asserting that appellants’ failure to wear seat belts was evidence of their comparative fault, appellee had the burden of proving their nonuse of seat belts was the proximate cause of some or all of their injuries. Appellee did not meet this burden. Except for the testimony of the officer who investigated the accident that appellants told him they were not wearing their seat belts, the only evidence offered on this issue came by way of appellants’ admissions on cross examination that they were not wearing their seat belts when the collision occurred. Their admissions alone are not sufficient proof that the nonuse of their seat belts proximately caused some or all of their injuries.
There was no testimony concerning the specific type of seat belt available to appellants and the effect of wearing a similar belt in a similar accident. There was testimony that appellant John Baker suffered injuries to his back and that appellant Carol Baker suffered injuries to her neck, shoulder, and knee. However, there was no testimony concerning the relationship of appellants’ nonuse of seat belts to these injuries. Without any evidence that their nonuse caused their injuries, the trial court erred in denying appellants’ motion in limine and allowing appellee to cross examine on this issue.
Although the jury was properly instructed on comparative fault, because this case was submitted to the jury as a general *463verdict, we are unable to determine how the jury apportioned fault or if it did so at all. The erroneously admitted evidence may very well have led the jury to attribute appellants with more fault than should have been the case. We therefore reverse and remand for a new trial.
Reversed and remanded.
Hays, J., dissents.
Glaze, J., not participating.