Guides · Any expert

Amended Rule 702 After Nearly Three Years: What Courts Exclude

Amended Federal Rule of Evidence 702 took effect on December 1, 2023. The Advisory Committee called it a clarification, and the appellate decisions largely agree, while saying more plainly than before that there is no presumption of admissibility and that the sufficiency of an expert's basis and the application of a method are questions for the judge. This guide covers what the circuits have done and what it means for preparing or attacking any expert, not only economists.

Guide 4 of 4 · September 29, 2026 · 8 min read
On this page
  1. What changed
  2. Preponderance
  3. Facts or data
  4. Reliable application
  5. The limit
  6. What to prepare

Circuit by circuit, with every appellate ruling quoted and linked: the Rule 702 tracker, reviewed monthly.

What the amendment changed

As quoted by the en banc Federal Circuit, the amended rule permits expert testimony "if the proponent demonstrates to the court that it is more likely than not that" four conditions are met, including that "(b) the testimony is based on sufficient facts or data" and "(d) the expert's opinion reflects a reliable application of the principles and methods to the facts of the case." EcoFactor, Inc. v. Google LLC, 137 F.4th 1333 (Fed. Cir. 2025) (en banc). The prior text of (d) asked only whether "the expert has reliably applied the principles and methods to the facts of the case." Id.

The courts have relied heavily on the Advisory Committee's note, which stated that "many courts have held that the critical questions of the sufficiency of an expert's basis, and the application of the expert's methodology, are questions of weight and not admissibility. These rulings are an incorrect application of Rules 702 and 104(a)." Id. (quoting the 2023 note). The same note says that "[n]othing in the amendment imposes any new, specific procedures," and, as the Sixth Circuit has quoted it, that "nothing in the amendment requires the court to nitpick an expert's opinion in order to reach a perfect expression of what the basis and methodology can support." Davis v. Sig Sauer, Inc., 126 F.4th 1213 (6th Cir. 2025).

A timing point: several appellate decisions reviewed rulings made under the prior text and said the result would be the same either way. Engilis v. Monsanto Co., No. 23-4201 (9th Cir. Aug. 12, 2025) ("our decision would be the same under either version of the Rule"). Because the circuits treat the amendment as restating existing law, the real target is pre-amendment language that treated basis and application questions as matters of weight.

Preponderance, and no presumption of admission

The Ninth Circuit's Engilis opinion is the clearest statement: "we confirm that a proponent of expert testimony must always establish the admissibility criteria of Rule 702 by a preponderance of the evidence and that there is no presumption in favor of admission." The court explained that, properly applied, the rule "requires that challenges to an expert's opinion go to the weight of the evidence only if a court first finds it more likely than not that an expert has a sufficient basis to support an opinion." Earlier circuit statements about a "liberal thrust" favoring admission "should not be understood to suggest a presumption of admission."

The Eighth Circuit rejected the argument that a district court overstepped by excluding an expert under the amended rule, holding that "after Rule 702's recent amendment courts continue to have a gatekeeping role to assure that evidence admitted in a case is both relevant and reliable." Sprafka v. Medical Device Business Services, Inc., 139 F.4th 656 (8th Cir. 2025).

Older formulations survive. In Davis v. Sig Sauer, the Sixth Circuit restated the preponderance burden and, in the same passage, quoted its precedent that "rejection of expert testimony is the exception, rather than the rule." Expect both lines of language in briefing.

Sufficient facts or data: the judge checks the foundation

The most consequential decision is EcoFactor. A patent damages expert testified that three lump-sum settlement licenses reflected the licensees' agreement to a specific per-unit royalty. The licenses' recitals stated only the patentee's belief about that rate, one license stated that the lump sum "is not based upon sales and does not reflect or constitute a royalty," and the expert had seen no licensee sales data. The en banc court held that this "fundamental premise" was "not based on sufficient facts or data, as required by Rule 702(b)," and ordered a new damages trial. It added: "Where, as here, the relevant evidence is contrary to a critical fact upon which the expert relied, the district court fails to fulfill its responsibility as gatekeeper by allowing the expert to testify at trial." Sufficiency is measured against the facts the expert actually relied on, not the whole record, and the court drew the line this way: "While the credibility of an expert's damages calculation is properly left to a jury, a determination of reliability under Rule 702 is an essential prerequisite."

The same principle governs medical causation. In Engilis, an oncologist's differential etiology ruled out obesity as a cause of the plaintiff's cancer based on a plaintiff fact sheet, although medical records showed a body mass index above 30 at two points in time. At the hearing the expert conceded he "certainly can't say whether [Engilis is] obese or not." The court held that "what matters is the evidence [the expert] actually considered and the conclusions he actually drew from that evidence in the process of forming his opinion as disclosed in his expert report." Explanations offered for the first time on cross-examination could not cure the gap, because Rule 26 requires the basis to be in the report.

Engineering opinions face the same scrutiny. In Sprafka, a design-defect opinion about a knee implant rested on case studies of fifteen and three patients that counted revisions without the number of devices implanted, plus a surgeon's impression of a high revision rate, while registry data suggested a similar or lower rate. The Eighth Circuit affirmed exclusion: without a reliable basis for the failure rate "or data showing how the ATTUNE system compares with other devices on the market, [the expert's] opinions lack reliability."

Litigation-generated data are not exempt. In Gilbert v. Lands' End, Inc., No. 23-3162 (7th Cir. Oct. 23, 2025), a causation expert relied entirely on questionnaires sent only to employees already suing, designed to streamline discovery, and returning erroneous responses, and he did not identify or control for those problems. Applying the pre-amendment rule, the Seventh Circuit affirmed exclusion and observed that the expert's burden "was clearly not an onerous one." In the court's words: "Less was required but not done here."

Reliable application: the method has to be followed

Amended subdivision (d) focuses on what the expert did with a sound method. In In re Onglyza (Saxagliptin) & Kombiglyze (Saxagliptin & Metformin) Products Liability Litigation, 93 F.4th 339 (6th Cir. 2024), the court accepted that "Bradford Hill is undeniably a reliable methodology," noted that the defense experts used it too, and still affirmed exclusion, because "the district court had an independent duty to ensure that all experts 'reliably applied' Bradford Hill." In Engilis, the district court excluded the opinion because the expert failed to follow the differential etiology his own report described.

Application is assessed opinion by opinion. In Davis v. Sig Sauer, two experts had expressly declined to investigate how the plaintiff's pistol discharged. The Sixth Circuit affirmed their exclusion on what caused the incident but reversed the exclusion of their opinions on design defect and reasonable alternative design, which were reliable and relevant on those elements. The note's instruction that "each expert opinion must stay within the bounds of what can be concluded from a reliable application of the expert's basis and methodology" cuts both ways: it trims overreach and protects opinions that stay inside the method.

The limit: gatekeeping is not weighing

The amendment does not license a judge to pick the more persuasive expert. Daubert itself identified "[v]igorous cross-examination, presentation of contrary evidence, and careful instruction on the burden of proof" as "the traditional and appropriate means of attacking shaky but admissible evidence." Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 596 (1993). Engilis adds that "'shaky' expert testimony, like any expert testimony, must still be 'admissible.'"

Rutledge v. Walgreen Co., No. 24-916 (2d Cir. July 13, 2026), shows the boundary. In the acetaminophen litigation, the district court excluded five general causation experts. The Second Circuit held that for three of them the court "went beyond its proper role as gatekeeper," including by excluding testimony consistent with the methodologies actually used in the field, while affirming exclusion of the other two. It described the 2023 amendment as a correction of a "hands-off" approach and accepted that a district court "must take a hard look at the experts' applications of accepted methods," but it directed courts to strike a middle ground, "taking a hard look at the expert's opinions to ensure that they are reliable without necessarily addressing whether they are correct." EcoFactor drew the same distinction from the other side, stressing that it was "not a case where the relevant evidence can reasonably support competing conclusions."

The practical line is between a premise that is unsupported or contradicted, which is the court's question, and a conclusion that reasonable experts dispute, which is the jury's.

What to prepare

  • List every factual premise each opinion depends on, and tie each one to material the expert actually relied on and cited.
  • Confirm that nothing in the record directly contradicts a critical premise.
  • Put the full basis in the Rule 26 report; the hearing is too late to supply it.
  • For each opinion, name the method and show that each step was followed.
  • Separate opinions so a weak one does not take down the others.
  • For any rate or comparison, make sure there is a denominator and a comparator.
  • When attacking, frame the challenge as basis or application, not disagreement with the conclusion.
  • Check whether the circuit's pre-2023 "weight, not admissibility" language still appears in recent opinions.

Supreme Mind builds Expert Witness Simulation for any expert on any matter, from case intake to settlement. A lawyer points it at a matter and picks a class of expert, and one run returns a brief in four sections: the expert's likely opinion, ranked cross-examination weaknesses, the Daubert / Rule 702 attack surface, and a directional settlement range. Experts are de-identified composites of a class, never a named individual; supporting quotations are checked verbatim against the public record, authorities carry their check result, and nothing it produces is filed or testifies. A complete sample brief is free, and the method explains how it works.

Sources

This guide is general information, not legal advice.

Run this analysis against the expert on your own matter.Start freeRead a sample brief →

The series

01Securities · September 8, 2026Daubert Challenges to Event Studies in Securities Cases02Securities · September 15, 2026Price Impact at Class Certification After Goldman03Antitrust · September 22, 2026Common Impact and Pass-Through at Class Certification04Any expert · September 29, 2026Amended Rule 702 After Nearly Three Years: What Courts Exclude

Related: the expert class library, a complete sample brief, and how it works.