Introduction
After Highway Patrol Trooper Michael Newton stopped Michael Nolte for a traffic violation on Interstate 70, Paul Daniel, driving a Trade Winds Distributing, Inc. truck, rammed the highway patrol car in which Newton and Nolte were sitting. The patrol car exploded. Trooper Newton died, and Nolte sustained serious injuries.
In the hotly contested trial of this case — with claims for wrongful death and serious personal injury arising from the manufacture of the Ford patrol vehicle— the trial court ruled that four prior incidents of gas tank explosions would be admissible to show that Ford had notice of explosions after it had remedied a defect in the car’s fuel system. The trial judge ruled inadmissible all evidence of explosions that occurred after the Newton accident. Despite this ruling, during trial, Ford in its own case presented evidence that, including the Newton explosion, there had been a total of 11 such incidents, four occurring before the explosion in this case and six occurring after the Newton explosion. In final argument, the trial court erroneously barred plaintiffs’ counsel from referring to the other explosion cases in evidence, mistakenly believing that, as per the trial court’s initial ruling on the issue, the accidents had been excluded. Ford’s attorney was able to capitalize on the error by arguing that the sole defect in the fuel system had been remedied. Due to the trial court’s erroneous ruling, plaintiffs were barred from using evidence of the six post-Newton accidents to rebut this argument.
The jury returned a verdict against plaintiffs and in favor of Ford and a verdict in favor of plaintiffs against Trade Winds Distributing, Inc., the employer of the driver of the pick-up truck that caused the collision by striking the patrol car. The jury’s verdict against Trade Winds awarded $4 million in damages to Newton and $4.5 million to the Noltes. The judgment against Trade Winds was not appealed and has become final.
In Newton’s and the Noltes’ post-trial motions on the verdict in favor of Ford, the trial court acknowledged its error during final argument but overruled the motion for new trial, holding that the error did not prejudice the plaintiffs.
The determination of prejudice rests largely within the discretion of the trial court as the referee at the scene of the contest. There are cases, however, where a replay of the scene, in an appellate court far removed from the heat of the contest, shows that the error was, indeed, prejudi*827cial. When evidence admitted during trial is excluded from being discussed in final argument, an appellate court presumes that the exclusion was prejudicial. This later appellate review also is influenced by whether the party benefiting from the trial court’s mistake leaves well enough alone or uses the mistake to its advantage. Here, Ford did not leave well enough alone. There does not appear to be a sufficient basis to rebut the presumption of prejudice. The trial court’s determination that plaintiffs did not suffer prejudice gives way to this Court’s conclusion that the error in excluding argument as to the evidence of the other explosions denied plaintiffs a fair trial against Ford.
The judgment in favor of Ford is reversed, and the case is remanded.
Facts and Background
While patrolling Interstate 70 on the morning of May 22, 2003, State Trooper Michael Newton stopped eastbound driver Michael Nolte for a minor traffic violation. Both vehicles pulled onto the shoulder of the highway, and Trooper Newton asked Nolte to accompany him to the patrol car so Newton could write up a warning for the traffic violation. While both men sat in the patrol car, Paul Daniel, the driver of the Trade Winds pick-up truck pulling an empty trailer, was traveling eastbound on 1-70. As he neared the patrol car, Daniel veered onto the shoulder of the interstate and collided with Trooper Newton’s vehicle. Upon impact, the patrol car burst into flames, killing Trooper Newton. Nolte survived but sustained very serious burns as a result of the explosion. Because neither man broke any bones in the collision, the evidence indicated that the injuries probably would not have been as serious if the fire had not occurred.
Plaintiffs Michael Nolte, his wife Barbie Nolte and Shonnie Newton, the widow of Trooper Newton, brought an action against Ford Motor Company, the manufacturer of the patrol car, and Trade Winds Distributing, Inc., Daniel’s employer. The claim against Trade Winds was for the negligence of its driver, Daniels. Plaintiffs asserted both negligence and strict products liability claims against Ford on the basis that the anti-spill valve and placement of the fuel tank behind the rear axle of the patrol car were defects that rendered the patrol car unreasonably dangerous. Plaintiffs argued that the unsafe placement of the fuel tank and the defective design of the anti-spill valve caused the explosion that killed Trooper Newton and injured Michael Nolte.
The anti-spill valve and the placement of the fuel tank in the Crown Victoria Police Interceptor
At trial, plaintiffs argued that a defect in the anti-spill valve of the patrol car, a Ford Crown Victoria Police Interceptor, coupled with the defective placement of the fuel tank behind the car’s rear axle caused the explosion that killed Trooper Newton and injured Nolte. The anti-spill valve consists of a spring and sealing flange and is designed pursuant to federal environmental regulations to prevent escape of vapors and liquid from the fuel tank. The anti-spill valve is located inside the fuel tank filler tube, the tube through which gasoline pumped at the filling station travels to the patrol car’s fuel tank. Plaintiffs argued at trial that, due to a defect, the anti-spill valve failed to prevent gasoline from escaping from the fuel tank when Trooper Newton’s patrol car was struck by Daniel’s truck and that, upon impact, gasoline spewed out of the patrol car’s fuel tank filler tube and ignited, resulting in the fatal explosion.
In addition to arguing that the anti-spill valve was defective, plaintiffs also argued *828that the placement of the patrol car’s fuel tank was an unreasonably dangerous defect. The fuel tank in the Crown Victoria patrol car is located behind the car’s rear axle, a location that, according to plaintiffs, places the fuel tank within the patrol car’s “crush zone,” the area likely to be crushed in an accident. As a result of this defective placement, according to plaintiffs, the force of the collision caused the neck of the patrol ear’s fuel tank filler tube to tear from the fuel tank itself, resulting in a significant leakage of gasoline at the severance site. Plaintiffs contended that this leakage, coupled with the leakage resulting from the failure of the anti-spill valve, caused the explosion in this case.
The shield upgrade kit
Both sides presented evidence at trial concerning the “shield upgrade kit” developed by Ford in 2002 for use on Crown Victoria Police Interceptor patrol cars. The shield upgrade kit was designed to cover components of the rear axle in order to prevent the shielded components from puncturing the fuel tank during rear-impact collisions. According to the testimony of Ford’s executives and expert witnesses, the shield upgrade kit was developed in response to field reports of fuel leakage resulting from puncture of the Crown Victoria Police Interceptor’s fuel tank during rear-impact collisions. Trooper Newton’s patrol car was equipped with the shield upgrade kit.
Evidence of other accidents involving Crown Victoria Police Interceptors equipped with the shield upgrade kit
Prior to trial, plaintiffs expressed their intention to introduce evidence that there had been 11 rear-impact collisions, including Trooper Newton’s accident, in which fuel leakage from Crown Victoria Police Interceptors equipped with the shield upgrade kit had resulted in fires. Of these 11 “post-upgrade” accidents, four preceded Trooper Newton’s accident. The other six post-upgrade accidents occurred after Trooper Newton’s collision.
At a pre-trial hearing on the issue, Ford objected to admission of the 11 post-upgrade accidents, arguing that they were not “of like character that occurred under substantially the same circumstances and resulted from the same cause” and were, therefore, inadmissible under Lopez v. Three Rivers Elec. Co-op., Inc., 26 S.W.3d 151, 159 (Mo. banc 2000). Plaintiffs argued that, even if not admissible as substantially similar accidents, evidence of the post-upgrade collisions with fuel leakage was admissible to show that Ford had notice of the continued danger of gasoline leakage during rear-impact collisions from Crown Victoria patrol cars equipped with shield upgrade kits.
The trial court agreed with plaintiffs, ruling that evidence of the other accidents was admissible to show Ford’s notice of danger caused by fuel leakage in patrol cars equipped with shield upgrade kits. Because only the accidents that occurred prior to Trooper Newton’s were relevant to show such notice, however, the trial court limited admission of evidence regarding other accidents to the four “pre-New-ton” collisions.
Discussion of the 11 post-upgrade accidents during trial
Despite the trial court’s ruling that only evidence of the four pre-Newton accidents was admissible, the fact that there had been a total of 11 post-upgrade accidents involving fuel leakage was presented twice at trial once in the deposition testimony of Ford’s vice-president of safety and once during plaintiffs’ cross-examination of one of Ford’s experts. Ford’s counsel first mentioned that there had been 11 post-*829upgrade accidents when counsel read to the jury the following portion of the deposition of Susan Cischke, Ford’s vice-president of safety:
Q: (By Ford’s counsel) There have been incidents with shields. It’s a fact. Everyone knows it. Are you aware of it?
A: Yes. Not only am I aware of it, but we talked about that at the press conference that there would continue to be— accidents would happen, and that we did not think that installation of the shields would eliminate all these types of accidents, and there would be indeed accidents that involved fuel leakage, eventually fire of vehicles that did have shields on them.
Q: Have you had a chance to look at some of the information that is available, that’s part of the court files here, concerning these incidents?
A: Yes, I have.
Q: And do you have some observations that you think might be helpful to understand from your perspective what your perspective is on some of these incidents?
A: Sure. I could share that with you. Overall, when I look, there’s about 11 accidents, I think, that are — are shown up there with the shield there that are vehicles that have been involved in a rear impact that had either some fuel leakage and some had fire. I know from our experience working with law enforcement that there’s probably many more out there that have had impacts with shields that had no leakage, and we know that from experience even during the whole processes. I mean, we had many letters from law enforcement agencies indicating, you know, for instance, in California, they total a vehicle a week and they haven’t had any instances of these. So while this represents vehicles that did have some leakage and some had fire, we know that there are many more others that — that did not have.
After reading this testimony into the record, Ford’s counsel approached the bench and informed the court that he inadvertently had allowed testimony as to all post-Newton incidents and that he did not wish to go further with the testimony. Ford did not request that the trial court strike the testimony, however, and plaintiffs did not object.
Later, the subject of the 11 post-upgrade accidents came up again when plaintiffs’ counsel referenced the 11 accidents during cross-examination of one of Ford’s experts. Ford did not object.
Closing Arguments
During closing arguments, plaintiffs’ counsel argued that the Crown Victoria Police Interceptor driven by Trooper Newton was unreasonably dangerous due to the placement of the fuel tank and a defect in the anti-spill valve and that these defects caused the fuel leakage in this case. Ford’s installation of the shield upgrade kit to prevent fuel leakage during rear-impact collisions was merely a “band-aid” measure that, plaintiffs argued, did not address the real underlying defect — the vulnerable placement of the fuel tank. In support of the argument that the shield upgrade kit did not effectively address the patrol car’s relevant defect, plaintiffs’ counsel attempted to discuss the 11 post-upgrade collisions with fuel leakage. Plaintiffs’ counsel stated,
“Ms. Cischke, the head of safety, said, ‘There have been 11 accidents with the shield that involved rear impacts that had some fuel leakage and some fire.’ After the shields were put on in the fall of 2000, [there were] 11 other accidents *830with fuel leakage and fire with the shields.”
Ford objected to plaintiffs’ mention of all 11 accidents, asserting that it referenced evidence that had been withdrawn and therefore was not in the case. The trial court sustained Ford’s objection and instructed the jury to disregard plaintiffs’ statements concerning the post-upgrade accidents. Forced to proceed without reference to the 11 accidents, plaintiffs’ counsel instead stated that the shield upgrade kits installed by Ford to prevent fuel leakage were effective only “in limited circumstances.”
During closing argument, Ford’s counsel argued that Trooper Newton’s patrol car was neither unreasonably dangerous nor defective. According to counsel, Ford had effectively dealt with the only defect in the Crown Victoria Police Interceptor — the risk of fuel tank puncture — by Ford’s installation of the shield upgrade kit in 2002. According to Ford, the fuel leakage and fire that caused Trooper Newton’s death were due to the high speed and impact of the collision and were not the result of any defect of the patrol car. In response to plaintiffs’ assertion that the shield upgrade kits cured the patrol car’s fuel leakage defect only “in limited circumstances,” Ford’s counsel stated:
Well, here are the limited circumstances that it works under, ladies and gentlemen. That upgrade kit and those countermeasures — you heard Dick Cup-ka and Jack Ridenour explain this to you — they were developed as a result of the joint efforts of law enforcement and Ford Motor Company that gathered information about every wreck that had occurred involving a police officer and, for that matter, a Panther. Every one.
They didn’t put their heads in the sand. They didn’t keep this from the police. Every piece of information about what happened in those incidents was gathered and collected, studied and analyzed, and that is how the upgrade kit was developed. And that was the product that was on this car when this accident occurred in May of 2008. Now, every one of the leakage modes, with the exception of the Lynn Ross incident, which nobody knows enough about to be able to say this with any certainty one way or the other, every one of those leakage modes was addressed by the upgrade kit. This is limited circumstances?
The jury’s verdict and post-trial proceedings
The jury returned a verdict in favor of plaintiffs on their claim against Trade Winds Distributing and awarded damages of $4 million to Newton and $4.5 million to the Noltes. On plaintiffs’ claim against Ford, the jury returned a verdict in favor of Ford, finding that Ford was not liable for the injuries sustained by plaintiffs as a result of the collision.
Plaintiffs filed a motion for a new trial, asserting that the trial court had abused its discretion by prohibiting plaintiffs from discussing all 11 accidents involving fuel leakage in patrol cars equipped with the shield upgrade kit. At the hearing on the motion, the trial court acknowledged that it had erred in prohibiting plaintiffs from referencing all 11 accidents because Ford had injected the issue of the six post-Newton accidents into the case when Ford’s counsel read Cischke’s deposition testimony into the record. Despite this error, the trial court determined that plaintiffs’ inability to discuss post-upgrade accidents had not been prejudicial and overruled plaintiffs’ motion for a new trial. On appeal, following opinion in the court of appeals, this Court granted transfer and *831has jurisdiction. Mo. Const, art. V, sec. 10.
Analysis
There is no question that the trial court erred in sustaining Ford’s objection to plaintiffs’ discussion of the 11 post-upgrade accidents during closing arguments. At trial, Ford’s counsel read deposition testimony in which Cischke stated that, since the installation of the shield upgrade kit in Crown Victoria Police Interceptors in 2002, there had been 11 rear-impact collisions involving upgraded patrol cars where fuel leakage was reported. By reading this testimony into the record, Ford’s counsel injected evidence of all 11 accidents into the case. Thus, once the accidents were in evidence, plaintiffs were entitled to discuss all 11 accidents despite the court’s earlier ruling that only the four “pre-Newton” accidents were admissible. As such, the trial court clearly erred in barring plaintiffs’ discussion of any of the 11 accidents during closing arguments.
With clear error established, the only question before this Court is that of prejudice — that is, whether plaintiffs were prejudiced by the trial court’s refusal to allow discussion of any of the post-upgrade accidents during closing arguments.
Was the trial court’s ruling prejudicial?
On review, this Court will consider a trial court error prejudicial if it appears from the record that the error “materially affected the merits of the action.” Rule 84.13(b); Lewis v. Wahl, 842 S.W.2d 82, 84-85 (Mo. banc 1992). This Court follows the analysis in Tune v. Synergy Gas Corp., 883 S.W.2d 10, 22 (Mo. banc 1994), holding that the trial court’s error in denying counsel the opportunity to discuss evidence in the case during closing argument is presumed prejudicial. The burden on Ford is to show that the error was not prejudicial.
Ford argues that plaintiffs’ inability to discuss the 11 post-upgrade accidents during closing arguments was not prejudicial because the effectiveness of the shield upgrade kit was not a material element to plaintiffs’ theory of liability. Ford points out that the purpose of the shield upgrade kit is to prevent fuel leakage by preventing the shielded components of the patrol car’s rear axle from puncturing the fuel tank. Since plaintiffs’ theory at trial was that the fire in this case resulted from a defect in the patrol car’s anti-spill valve and defective placement of the fuel tank — and not a fuel tank puncture of the sort that the shield upgrade kit was designed to prevent — Ford argues that the effectiveness of the shield upgrade kit was not an element material to the jury’s determination of liability. As such, Ford argues that plaintiffs’ inability to discuss the post-upgrades accidents could not have been prejudicial.
Review of the record, however, demon-' strates that the impact of the trial court’s error was not as immaterial to the jury’s determination of liability as Ford claims. Throughout the trial, Ford argued that the shield upgrade kit addressed the only defect in the patrol car’s fuel tank. According to Ford, the gasoline leakage that caused Trooper Newton’s patrol car to catch fire was not the result of any defect but rather was the unpreventable result of the collision’s powerful impact. In support of this argument, Ford presented testimony of witnesses, including Cischke, who stated that the shield upgrade kit addressed the only defect in the patrol car’s fuel tank that could have resulted in fuel leakage.
To counter this argument, plaintiffs wished to argue during closing that the evidence of the 11 post-upgrade accidents *832involving fuel leakage supported a contrary conclusion — that the shield upgrade kit did not address the patrol car’s underlying defect as Ford claimed. If the jury believed plaintiffs’ argument that the post-upgrade accidents indicated the existence of a continuing defect — a permissible inference from the evidence — this argument would have refuted Ford’s claim that the fuel leakage and fire in this case were the unavoidable results of circumstance. Because the trial court refused to allow plaintiffs’ discussion of the 11 post-upgrade accidents during closing argument, however, plaintiffs were barred unfairly from presenting this theory to the jury.
Ford points out that, although barred from discussing all 11 of the post-upgrade accidents, plaintiffs were able to present evidence regarding the four post-upgrade collisions that occurred before the Newton accident. In the context of the evidence, the logic of this argument is tenuous. Ford’s contention is that each of these accidents is an isolated and random incident. That explanation is plausible when it pertains to only four accidents. Eleven accidents may strain credulity.
Ford aggravated the unfair impact of the trial court’s ruling by taking advantage of the error during its own closing argument. During closing, Ford argued to the jury that “every one of the leakage modes ... was addressed by the upgrade kit.” In other words, Ford argued that the only defect that could have caused gasoline leakage was cured by the shield upgrade kit. Because the trial court refused to allow plaintiffs to discuss the post-upgrade accidents, plaintiffs were unable to rebut this argument. Ford’s contention that the shield upgrade kit effectively addressed the only leakage defect went uncontested.
As Ford’s emphasis on the issue during closing indicates, the effectiveness of the shield upgrade kit was an important issue in this case, even though Ford now argues that it was collateral. As a result of the court’s refusal to allow plaintiffs to discuss the post-upgrade collisions during closing, plaintiffs were unable to present their theory regarding this issue to the jury. Because Ford was able to present its argument during closing that any defect in the patrol car’s fuel tank was cured by the upgrade, this ruling put plaintiffs at an unfair disadvantage. Despite the fact that the subject of this imbalance (the effectiveness of the shield upgrade kit) is not a specific element of plaintiffs’ claim, the effectiveness of the upgrade nevertheless seems significant to the jury’s determination of liability. The highly technical subject matter of this case increases the likelihood that any imbalance in the parties’ ability to present and argue inferences from evidence would influence the jury’s verdict.
Based on the unfairness of the trial court’s error, this Court finds that the trial court’s refusal to allow plaintiffs’ to discuss the post-upgrade accidents during closing arguments was an error “materially affecting the merits of the action.” Rule 84.13(b). As such, the trial court abused it discretion in overruling plaintiffs’ motion for a new trial.
Conclusion
The judgment of the trial court is reversed, and the case is remanded.
STITH, C.J., PRICE, TEITELMAN and FISCHER, JJ., and DOWD and McELWAIN, Sp.JJ., concur.
WOLFF, J., concurs in separate opinion filed.
PRICE, J., and McELWAIN, Sp.J., concur in opinion of WOLFF, J.
RUSSELL and BRECKENRIDGE, JJ., not participating.