WELLFORD, J., delivered the opinion of the court, in which MERRITT, C.J., joined. KEITH, J. (pp. 226-27), delivered a separate opinion concurring in part and dissenting in part.
WELLFORD, Circuit Judge.Plaintiff, Gary Wayne Brock, was a longtime bulldozer operator for Nally & Hamilton (“N & H”), a reclamation company doing work in coal mining areas in mountainous Harlan County, Kentucky. In June of 1990, Brock began work on a “high wall” where strip mining had taken place.1 Brock normally operated a Caterpillar model D8L bulldozer, but on one particular day, N & H assigned him to a larger model D9H, which was also manufactured by defendant Cater*222pillar, Inc. (“Caterpillar”). Great Western Coal Company (“Great Western”) owned the bulldozer, but N & H had borrowed it because it was already at the work site. Brock conceded in his testimony that he had never before operated a model D9H Caterpillar bulldozer.
Brock testified that he first checked the bulldozer’s fluid levels before commencing work.2 In order to reach the top of the fill, Brock was required to push dirt up an access road near the top of the mountain. The slope was approximately a two-to-one grade, an angle of about twenty-seven degrees. As Brock proceeded up the access road, he testified that the bulldozer suddenly lost “its prime” or its hydraulic fluid pressure, which caused the bulldozer to lose its braking power on this slope. Brock attempted to slow the precipitate backward fall by dragging the bulldozer’s blade unsuccessfully and attempting manually to apply the brakes. As the bulldozer spun out of control, it left the access road, twisted sideways, and tumbled down the reclaimed high wall, throwing Brock from the cab of the vehicle. The D9H continued its fall, eventually coming to rest ■with the engine still running. After the accident, N & H employees retrieved the bulldozer and found it to be functioning normally. They drove the D9H back up the same access road, and this bulldozer continued in operation for several years to the date of trial.
In the accident, Brock suffered injuries to his cervical spine, ankle, lower back, and his psyche. He had an operation on his ankle and another has been recommended with respect to claimed ongoing neck problems. As a result of the accident, physicians have recommended that Brock seek counseling and psychiatric care for the depression he has allegedly experienced subsequent to the accident. He has not, since the accident, operated a bulldozer.
In June, 1991, Brock, a resident of Kentucky, brought a timely diversity action against Caterpillar only, alleging that Caterpillar defectively designed the D9H bulldozer and that the defective design made it unreasonably dangerous and was the proximate cause of his injuries.3 Brock alleged that in 1978, around the time that this D9H was manufactured and sold, Caterpillar had designed a safer braking system for bulldozers which caused disc brakes to be applied automatically whenever hydraulic pressure was lost. That improved system was not ever used in this model D9H. Rather, Caterpillar elected to manufacture the D9H with a “hydraulic boosted” or “band” braking system, which the bulldozer operator applies manually through means of a brake pedal, despite this claimed availability or technical knowledge of particular safer braking mechanism. The “hydraulic boosted” system supplied the extra hydraulic power under normal operating conditions, and served as a supplementary source of power for the brakes on the model D9H.
Brock contended that the D9H “cativat-ed”4 as he moved up the steep incline and that the resulting loss of hydraulic pressure caused his brakes to fail. Brock further maintained that if Caterpillar had manufactured the D9H with the safer disc braking system, which applies the brakes automatically upon the loss of hydraulic pressure, then the bulldozer would not have “free wheeled” down the high wall.
After Brock filed his complaint, Old Republic Insurance Company (“Old Republic”), the worker’s compensation carrier for N & H, intervened, asserting a claim for recovery of medical expenses and other benefits already dispersed to Brock. The trustee for the worker’s compensation Special Fund also intervened as a plaintiff, seeking the recov*223ery of any benefits paid. Caterpillar brought a third-party claim against Great Western, claiming that if the accident occurred as Brock alleged, Great Western was responsible because it failed to observe routine care and maintenance of the bulldozer. Great Western, however, was dismissed on the basis of its post-bankruptcy discharge and stay. The parties agreed that, should the evidence warrant, Great Western might be included in the jury apportionment of any fault or liability.5 The two intervening plaintiffs did not actively participate in the trial nor are they participating in this appeal.
Prior to trial, Caterpillar moved to exclude evidence concerning the differences between the braking systems on the D9H and later model bulldozers, the D9L and the much larger DIO. Caterpillar argued that the models were so dissimilar that evidence regarding any comparison would be overly prejudicial and misleading to the jury. The magistrate judge denied defendant’s motion in limine, and the case proceeded to a jury trial.
The jury returned a verdict in favor of Brock, apportioning fault as follows: Caterpillar 60%, Great Western 20%, N & H 15%, and Brock 5%. The jury awarded Brock a total of $950,000.00, apportioned as follows: $20,000.00 for past medical expenses, $50,-000.00 for future medical expenses, $100,-000.00 for lost earnings, $630,000.00 for permanent impairment of Brock’s future earning capacity, $100,000.00 for past physical and mental suffering, and $50,000.00 for future mental suffering. Caterpillar moved for judgment as a matter of law or, in the alternative, for a new trial, arguing that there was no proper and relevant evidence from which a reasonable juror could find that Caterpillar defectively designed the D9H in question. The magistrate judge denied that motion. Caterpillar appeals that denial, and Brock cross-appeals the magistrate judge’s holding that there was insufficient evidence to support the jury’s award of damages for future medical expenses.
We address first Caterpillar’s argument that, because the jury found that Great Western failed “to observe routine care and maintenance” of the D9H, it was entitled to a judgment as a matter of law pursuant to Ky.Rev.Stat. AnN. § 411.320(1). Brock argues that Kentucky’s Comparative Fault Act, codified at Ky.Rev.Stat. Ann. § 411.182, repealed § 411.320(1) by implication. Because Kentucky has a substantial interest in the resolution of that question, we certified the following issue to the Kentucky Supreme Court in May of 1995:
Did Ky.Rev.Stat. Ann. § 411.182(1) negate and/or overrule Ky.Rev.Stat. Ann. § 411.320(1)?
In response, the court unequivocally pronounced that “KRS 411.182(1) has repealed KRS 411.320(1).” Therefore, we have no choice but to reject Caterpillar’s argument that § 411.320(1) provides it a complete defense in this action. Nevertheless, we find in favor of Caterpillar for the reasons that follow.
As is stated above, the tractor involved in Brock’s accident was a model D9H bulldozer, which was manufactured by Caterpillar and sold by its dealer some twelve years before the 1990 accident that brought about Brock’s injury. Apparently, this bulldozer had been utilized for many years without any problems of the nature asserted by Brock. Brock alleged, among other things, that Caterpillar “did test the vehicle for safety,” but that “it was defectively designed [and] unreasonably dangerous.” Caterpillar’s principal defense was that, as a matter of law, plaintiff failed to produce sufficient admissible evidence to show that the bulldozer’s braking system was defective due to its design or that it was unreasonably dangerous.6
*224In establishing a defect in product design, a plaintiff must show something more than that it was “theoretically probable that a different design would have been feasible.” Ingersoll-Rand Co. v. Rice, 775 S.W.2d 924, 928 (Ky.Ct.App.1988). Under applicable Kentucky law in this diversity case, plaintiff must also demonstrate “more than that a particular injury would not have occurred had the product which caused the injury been designed differently.” Jones v. Hutchinson Mfg., Inc., 502 S.W.2d 66, 70 (Ky.1973) (quoting 63 Am.Jur.2d Products Liability § 73, p. 79); Rice, 775 S.W.2d at 928-29 (citing Jones, 502 S.W.2d at 70). Another court, interpreting Kentucky law on this subject, stated that “proof that technology existed, which if implemented would feasibly have avoided a dangerous condition, does not alone establish a defect.” Sexton v. Bell Helmets, Inc., 926 F.2d 331, 336 (4th Cir.), cert. denied, 502 U.S. 820, 112 S.Ct. 79, 116 L.Ed.2d 52 (1991).
Like many other state courts, Kentucky courts have encountered difficulty with the meaning of the strict liability concept, “defective condition unreasonably dangerous.” Indeed, in Montgomery Elevator Co. v. McCullough, 676 S.W.2d 776, 780 (Ky.1984), the court stated that for fifteen years “we struggled with the meaning of these words.” McCullough pointed to Nichols v. Union Underwear Co., 602 S.W.2d 429 (Ky.1980), in interpreting § 402A of the Restatement (Second) of Toets, which had been adopted in Kentucky:
[ T]he question is whether the product creates “such a risk” of an accident of the general nature of the one in question “that an ordinarily prudent company engaged in the manufacture” of such a product “would not have put it on the market.”
McCullough, 676 S.W.2d at 780 (citing Nichols, 602 S.W.2d at 433).
The McCullough court pointed out pertinent considerations in these products liability eases to take into account: (1) “feasibility of making a safer product,” (2) “pa-tency of the danger,” (3) “warnings and instructions,” (4) “subsequent maintenance and repair,” (5) “misuse,” and (6) “inherently unsafe characteristics.” Id. The McCullough court held that evidence of product failures of “substantial similarity” was “relevant and admissible.” Id. at 783. Likewise, evidence of a lack of substantially similar instances of malfunction would be admissible for certain purposes. See Harris v. Thompson, 497 S.W.2d 422, 429 (Ky.1973) (“Evidence of the occurrence or nonoccurrence of other accidents or injuries under substantially similar circumstances is admissible when relevant to certain limited issues.... ”). The magistrate judge permitted Brock’s expert, William M. Cooper, to testify about the feasibility of using the improved brake system in the D9H used by Brock. The question on this appeal, then, becomes whether the admission of the “comparison” evidence regarding the use of another braking system by Caterpillar on a different model was relevant and proper, and whether Cooper’s opinion, thus, was properly admitted.
The evidence showed that Caterpillar made the model D9H bulldozer from 1974 to 1981, and produced almost 10,000 of such machines, which weighed about 47 tons, with “band type” brakes on each side, boosted by a hydraulic system. Cooper testified as follows:
Q. Do you have an opinion, sir, as to what effect if any the design of this bulldozer had on this accident?
A. Yes, if the machine had had a fail-safe brake system, such as the newer style tractor did, the D 10 in ’78 and the newer styled D 9’s did in I think ’80 or ’81, and therefore, if the machine had lost pressure it would have locked up and Mr. Brock wouldn’t have gone down the hill.
(emphasis added). , Cooper’s testimony that the D9H was defective and unreasonably dangerous was based upon a comparison with the D9H brake system and that incorporated in the newer and larger model DIO.7 In addi*225tion, he was allowed to base his opinion, improperly we think, on Caterpillar’s incorporation of the new and different disc type brake system, coupled with a new locking feature, in D9 models two or three years later. Cooper, was of the view that the machine lost hydraulic pressure causing a loss of boost to the brakes, but he could not indicate the reason for such failure. He admitted that inspection and testing of the machine after the accident revealed no physical evidence of loss of hydraulic pressure. Great Western, the owner of the equipment at that time, inspected the bulldozer and found nothing essentially wrong with the machine’s mechanism. In fact, with minor repairs, the machine was used without adverse consequences for three years after the accident.
The model D9H in dispute had already been manufactured for some years before the new system was first incorporated in the concededly much heavier, much more powerful higher capacity model DIO. We find no direct evidence that the equipment operated by Brock, particularly its brake system, was inherently dangerous. Cooper himself, familiar with these bulldozers, stated that he had previously considered them to be reasonably safe “at the time.” Moreover, even by 1980, Cooper “had no idea that there was anything else available.” In 1978, the D9H was concededly not “in violation of any applicable industry standards.” There was no evidence that any competitor of comparable machines in 1978 produced them with disc type or so-called “fail safe” brakes.8 Cooper had never worked in design of bulldozer equipment, and he worked in a family owned coal company which utilized similar Caterpillar D9 equipment. Cooper was familiar with D9H operating instructions which warned about services and maintaining for oil pressure if a bulldozer were operating on steeper slopes with recommendation of a three and half gallon overfill. Brock testified that his oil level “was a little over full,” and he checked for leaks before driving this bulldozer for the first time.
There were a large number of changes in character, size, style, and technological advancement between the manufacture design of the D9H in the early 1970’s and that of the DIO some years later. We are persuaded that the magistrate judge erroneously permitted Brock’s expert to compare these two models and their braking systems in concluding that the earlier D9H model at issue was defectively designed and unreasonably dangerous. It was analogous, in our view, to allowing a plaintiffs expert to compare unfavorably the braking system on a 1974 Chevrolet made by General Motors with that company’s 1978 Cadillac automobile. We, therefore, conclude that the magistrate judge committed prejudicial error in allowing this analogy to be drawn by plaintiffs expert as proof of defective and unreasonable dangerous design.
Nor, are we persuaded by Ford Motor Co. v. Fulkerson, 812 S.W.2d 119 (Ky.1991), to the contrary. A divided court in that case held that evidence of a change in the same model Ford vehicle from the year 1984 to 1985 was admissible in a design defect case, holding, however, that the trial judge retained “power to consider whether evidence of little probative value ‘may be more prejudicial than it would be helpful.’ ” Id. at 126.9 Under the circumstances of the instant case, we are satisfied that plaintiffs proof of later changes in another model should have been considered in the light of Fed.R.Evid. 403. In our view, although this “comparison” evidence may be tenuously relevant, such relevance is outweighed by the prejudicial effect in this case because of the dissimilarities in the bulldozers that were the subject of the comparison.10
*226Furthermore, Cooper conceded that he could not tell what caused the loss of hydraulic pressure to the brakes in question, and he could not attribute this situation to a design defect. He also conceded that with regard to oil pressure, the bulldozer is designed “so that the brakes are taken care of first.” (There was no evidence of malfunction of the priority valve.) At the time of the accident, moreover, Cooper considered this model Caterpillar used by Brock to be a “reasonably safe machine.” After the accident, it is not denied that everything about the tractor was “normal,” and it was driven back up the same, or a steeper, incline. Smaller Caterpillar tractors had no brake boosting system comparable to the D9H in question.
In short, Cooper was qualified to testify, and did testify, about matters pertinent to the condition and design of the braking system on the D9H machine at the pertinent time. We find error, however, in allowing him to testify as to the comparison between Caterpillar bulldozers that were were substantially different or marketed and sold at a considerably later time. It was noted in Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579, 113 S.Ct. 2786,125 L.Ed.2d 469 (1993), that “[e]xpert evidence can be both powerful and quite misleading.... Because of this risk, the judge in weighing possible prejudice against probative force under Rule 403 of the present rules exercises more control over experts than over lay witnesses.” Daubert, 509 U.S. at 595, 113 S.Ct. at 2798 (quoting Weinstein, 138 F.R.D., at 632). Daubert also pointed to an earlier ease from this circuit “holding that scientific evidence that provided foundation for expert testimony viewed in the light most favorable to plaintiffs, was not sufficient” to let the jury decide causation by the alleged defective product. Id. (citing Turpin v. Merrell Dow, 959 F.2d 1349 (6th Cir.), cert. denied, 506 U.S. 826, 113 S.Ct. 84, 121 L.Ed.2d 47 (1992) (emphasis added)). In the instant ease, relevant and admissible evidence was insufficient to establish a defective product. In the alternative, in light of defendant’s objections to admission of such testimony, the magistrate judge should have ruled it inadmissible under Fed.R.Evid. 403.
Other than this comparison proof presented by Cooper, which we have deemed inadmissible, plaintiff has failed to demonstrate a design defect under applicable Kentucky law.11 Accordingly, we must REVERSE the verdict for the plaintiff for insufficient proof of a brake system design defect on the bulldozer involved in this unfortunate accident.12
For the reasons indicated, we REVERSE the judgment for the plaintiff.
. A "high wall” is the vertical wall that remains on a mountain slope after coal has been stripped away. At the site in question, the fill had slipped at the top of the high wall and it was plaintiff's job to push the dirt back to the top.
. Caterpillar contends that an inspection subsequent to the accident indicated that the fluid levels were full, but not overfilled, as recommended in Caterpillar’s sales literature.
. The parties agreed to trial of the case by Magistrate Judge James F. Cook.
.Cativation is a term used to describe the drawing away of fluid from the intake valve to the hydraulic pump. This condition causes air to be pumped through the engine instead of oil or fluid. Cativation can result in the loss of hydraulic pressure when a bulldozer is operating on a steep slope. Caterpillar's manual recommended overfilling the bulldozer with hydraulic fluid to prevent cativation.
. The magistrate judge also ordered N & H, Brock’s employer, to be included in the jury's apportionment of fault.
. The only part of the bulldozer alleged to be defective at trial was the braking mechanism. However, Brock also alleged that “use of coal company employees ... was foreseeable and would likely result in severe injury.” He also averred that "[i]t was feasible and practical” at the time of manufacture to design the bulldozer so that it was safe for users. The problems specified were "inherent instability and handling difficulties ... [and] propensity to go out of gear.” Thus, the bulldozer was alleged to be violative of the condition required under the Kentucky Products Liability Act.
. Caterpillar began manufacturing a 90 ton bulldozer, the model DIO, in 1978. It is conceded that the model D10 was the first one to utilize the new brake system during 1978. There is no question but that the D9H was in its design stages some years before the newer, different system was first offered in the different, much larger, model D10 by Caterpillar.
. Only Cooper described the new system, made for the much heavier DIO equipment, to be "fail safe.”
. In that case, “Ford did not undertake to confront plaintiff's expert on the feasibility of additional retention features” (on the 1984 spindle design). Id. at 126.
.The D9H bulldozer had a six cylinder engine; the DIO bulldozer had a twelve cylinder engine. Caterpillar’s engineer, Buie, testified that there was "no way physically to put the DIO braking system into the D9H tractor.” The new D9L model made in 1981 weighed more than 57-1/2 tons (more than 20% heavier than the D9H model).
. Under all of these circumstances, the jury's allocation of sixty percent of the damages in this case to defendant Caterpillar was simply without reasonable basis. Even if we had found the evidence to be sufficient to find defendant's brake design to be a contributing factor to Brock's injury, we would have remanded the case for a reallocation of fault, including that attributable to plaintiff Brock.
. In light of this holding, we need not address Brock's argument that he is entitled to damages for future medical expenses.