J.B. Hunt Transport, Inc. (J.B. Hunt) appeals from the judgment of the district court, ^[1] entered upon a jury verdict in favor[*443] of defendants, General Motors Corporation (General Motors) and Fisher & Company (Fisher), in this products liability contribution action. J.B. Hunt asserts that the district court erred by excluding certain expert testimony and other offered evidence, and as a result, a new trial is warranted. We affirm.

I.

This appeal stems from a multiple vehicle collision occurring on February 19, 1994, on Interstate 70 (1-70) in St. Charles, Missouri. Richard and Alicia Spitzenberg were traveling westbound in a 1991 Chevrolet Camaro. Alicia was driving, and Richard was sitting in the passenger seat not wearing his seatbelt. As Alicia merged onto 1-70, a Toyota Corolla in front of her slowed, causing Alicia to decelerate and a J.B. Hunt 18-wheel tractor trailer to hit the rear end of the Spitzen-bergs’ Camaro. The initial impact caused Richard to lose consciousness and the Ca-maro to careen into the Toyota and a nearby Ford Crown Victoria. In turn, the Camaro spun and again collided with the J.B. Hunt truck. The Camaro finally came to rest facing northeast in the westbound lanes of 1-70.

Richard Spitzenberg filed suit against J.B. Hunt, asserting he sustained catastrophic injuries when his head hit the Camaro’s rear seatback during the accident. J.B. Hunt settled the suit with Richard for over $2.6 million and sued General Motors, the Camaro’s manufacturer, and Fisher, a component seat manufacturer, for contribution. J.B. Hunt proceeded to trial on the theory that the Camaro’s front seat was not crashworthy. J.B. Hunt contended that upon impact, the front seatback reclined, which allowed Richard to ramp up the reclined front seatback and strike the back of the rear seat, and that this sequence of events precipitated his injuries.

Prior to trial, defendants filed a motion in limine to exclude the testimony of J.B. Hunt’s accident reconstructionist, Jerry Wallingford. Defendants argued that Wallingford’s trial testimony would advance a theory not supported by the evidence; specifically, that the J.B. Hunt truck collided with the Camaro only once, making for a total of three impacts sustained by the Camaro — one impact with the truck, one with the Toyota, and one with the Ford. The district court denied defendants’ motion under the impression that J.B. Hunt would present eyewitness testimony supporting Wallingford’s three-impact theory.

During the jury trial, the district court allowed Wallingford to testify over defendants’ renewed objections. Wallingford testified that because he had insufficient information, he was unable to scientifically reconstruct the accident. (Tr. at 293, 309-10.) After hearing this concession, the district court excused the jury and independently heard the remainder of Walling-ford’s testimony. Wallingford stated to the court that, based upon his examination of photographs taken of the involved vehicles, the tractor trailer struck the Camaro only once, thus lending credence to his three-impact theory. The district court ruled that because uncontradicted eye witness testimony established that the tractor trailer hit the Camaro twice, it was excluding Wallingford’s testimony. The district court further concluded that the testimony was inadmissible in light of Daubert v. Metrell Dow Pharm., Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993), and that Wallingford’s photograph analysis as a proffered explanation for the three-impact theory lacked scientific support.

J.B. Hunt also called Anthony Sanees, a biomechanical expert and claimed expert in “foamology,” to testify about the manner in which Richard Spitzenberg received his injuries during the crash. The district court permitted Sanees to testify as to where he believed Richard made contact with the rear seat but refused to allow Sanees to bolster his testimony with any' analysis of the Camaro’s rear seat foam thickness. The district court stated:

[*444] He is not an expert in foam, and as best as I can tell, there is no science of foam.... [H]e has no idea what happens to foam under normal circumstances, so how can he have any scientific basis for basing any of his conclusions on what happened to the foam in this case?

(Tr. at 840.) Moreover, the district court concluded that Sances’s testimony was derivative of the already excluded Walling-ford testimony and therefore inappropriate.

At the conclusion of the eight-day trial, the jury found in favor of defendants. J.B. Hunt now appeals and requests a new trial.

II.

We review the district court’s decision to exclude expert testimony under Daubert for an abuse of discretion. Kumho Tire Co. v. Carmichael, 526 U.S. 137, 152, 119 S.Ct. 1167, 143 L.Ed.2d 238 (1999). Under Federal Rule of Evidence 702, expert testimony is admissible when it will assist the trier of fact. ^[2] See also Daubert, 509 U.S. at 597, 113 S.Ct. 2786 (concluding that the germane inquiry when determining admissibility of expert testimony is whether the testimony would assist the jury’s understanding of the evidence). Daubert makes clear that the district court must “ensur[e] that an expert’s testimony both rests on a reliable foundation and is relevant to the task at hand.” 509 U.S. at 591, 113 S.Ct. 2786.

We agree with the district court’s assessment that Wallingford’s proffered expert testimony fails to qualify as scientifically valid. Unlike defendants’ accident reconstructionist, whose testimony utilized the testimony given by other witnesses, Wallingford’s three-impact theory was premised primarily upon his impressions of the photographs of the scratches in the paint of the vehicles involved in the accident. In fact, Wallingford conceded he had insufficient evidence to completely reconstruct the accident as he theorized.

[N]othing in either Daubert or the Federal Rules of Evidence requires a district court to admit opinion evidence that is connected to existing data only by the ipse dixit of the expert. A court may conclude that there is simply too great an analytical gap between the data and the opinion proffered.

General Elec. Co. v. Joiner, 522 U.S. 136, 146, 118 S.Ct. 512, 139 L.Ed.2d 508 (1997). Because of the deficiencies at the core of his opinion, including his own admission concerning his inability to scientifically reconstruct the accident, Wallingford’s resulting conclusion — that this indeed was a three-impact collision — was mere speculation and pure conjecture. “Expert testimony that is speculative is not competent proof and contributes nothing to a legally sufficient evidentiary basis.” Concord Boat Corp. v. Brunswick Corp., 207 F.3d 1039, 1057 (8th Cir.) (internal quotations omitted), cert. denied, — U.S. -, 121 S.Ct. 428, 148 L.Ed.2d 436 (2000). The district court did not abuse its discretion by excluding Wallingford’s testimony.

Given that we find that the district court committed no error by excluding Wallingford’s testimony, we also find no error in excluding Sances’s testimony because the two are inextricably linked. Sanees offered testimony in order to explain how Richard Spitzenberg’s body would respond during a three-impact collision. Thus, Sances’s testimony had no evidentiary basis absent Wallingford’s testimony presenting the three-impact theory, and we therefore conclude the court committed no abuse of discretion in excluding it. Furthermore, we are skeptical of J.B. Hunt’s attempt to offer Sanees as an expert in “foamology.” Sances’s experience with analyzing foam stemmed purely from observations of its thickness. He received[*445] no formal training or course work in foam (nor do we know of any that exists in this area). The testimony submitted by Sanees was not derived from the application of any reliable methodology or scientific principle. It is well within the district court’s “discretion to choose among reasonable means of excluding expertise that is fausse and science that is junky.” Kumho, 526 U.S. at 159, 119 S.Ct. 1167 (Scalia, J., concurring).

J.B. Hunt further argues that the district court improperly excluded evidence of Alicia Spitzenberg’s minor injuries. J.B. Hunt contends that Alicia’s injuries were dispositive proof that Richard Spitzenberg’s seat failed given that both occupants were subjected to the same impact. In a products liability action, evidence of other injuries may be relevant to prove a product’s lack of safety or a party’s notice of defects. Lovett v. Union. Pac. R.R., 201 F.3d 1074, 1081 (8th Cir.2000). “However, admitting similar-incident evidence also threatens to raise extraneous controversial issues, confuse the issues, and be more prejudicial than probative.” Id. For these reasons, the party seeking admission of the evidence must demonstrate that the circumstances between the two incidents are substantially similar. Drabik v. Stanley-Bostitch, Inc., 997 F.2d 496, 508 (8th Cir.1993).

We cannot say that Alicia Spitzenberg’s injuries could be considered either substantially similar evidence or relevant to Richard Spitzenberg’s injuries. First, it was undisputed that Alicia, unlike Richard, was wearing her seatbelt at the time of the accident. Second, because Alicia was driving, she was able to anchor herself by grasping the steering wheel, which may have alleviated some of the impact. Moreover, the angle of impact differed between the two occupants leading to the varying degrees of injury. In our view, the district court correctly excluded the evidence of Alicia’s injuries.

J.B. Hunt also attempted to introduce evidence of tests conducted by General Motors in the 1960s on experimental, nonproduction seats to illustrate General Motors’ knowledge of the potential hazards associated with greater yielding seats. The district court found that the conditions of the tests were too dissimilar to the present accident to be admissible; moreover, the evidence was untimely and irrelevant to the dispute at hand. Admissibility of experimental tests is within the purview of the trial judge’s discretion. Bunting v. Sea Ray, Inc., 99 F.3d 887, 892 (8th Cir.1996). “A court may properly admit experimental evidence if the tests were conducted under conditions substantially similar to the actual conditions.” McKnight v. Johnson Controls, Inc., 36 F.3d 1396, 1401 (8th Cir.1994) (internal quotations omitted). We agree with the district court that these tests conducted thirty years ago on experimental seats would be both prejudicial to defendants and confusing to the jury. The district court correctly excluded this evidence.

Finally, J.B. Hunt contends that the district court should have admitted evidence of General Motors’ subsequent remedial measures regarding its current seat integrity. Under Federal Rule of Evidence 407, a company’s postaccident design change is inadmissible to prove fault — absent some exception. J.B. Hunt asserts that General Motors abrogated this evidentiary exclusion by disputing feasability. We conclude J.B. Hunt’s argument is without merit given that General Motors stipulated that stiffer seats were feasible. ^[3]

III.

Accordingly, we affirm the judgment of the district court.