Summary. On December 1, 2023, Federal Rule of Evidence 702 was amended to correct what the Advisory Committee described as widespread misapplication by courts. Many decisions had held that questions about an expert's application of a methodology go to weight rather than admissibility, effectively surrendering the gatekeeping role the Supreme Court assigned in Daubert. The amended rule makes two things explicit: the proponent must demonstrate by a preponderance of the evidence that each admissibility requirement is met, and the expert's opinion must reflect a reliable application of the methodology to the facts of the case. This article explains what changed and what did not, works through the Daubert trilogy and the reliability factors courts actually use, and covers the disclosure machinery of Rule 26(a)(2) including the retained/non-retained distinction and the work product protections added in 2010. It addresses the practicalities that decide expert fights: report drafting, deposition strategy, motions to exclude and their timing relative to summary judgment, the special problems of damages experts and apportionment, and cross-examination technique. It closes with checklists, a worked example, an FAQ, and related reading.

Expert testimony is the only kind of evidence in an American courtroom where a witness may tell the jury what to conclude. A fact witness says what they saw. An expert says what it means. That power is why the rules governing experts are more demanding than the rules governing anyone else, and why the fight over whether an expert may testify at all is frequently the most important motion in a case.

For twenty years after Daubert, that fight tilted steadily toward admission. Courts developed a comfortable formulation: challenges to an expert's methodology go to admissibility, but challenges to the expert's application of the methodology go to weight, to be explored on cross-examination. The formulation was tidy, quotable, and, according to the Advisory Committee on Evidence Rules, wrong.

The December 2023 amendment says so.

The short answer

Rule 702, as amended effective December 1, 2023, reads:

A witness who is qualified as an expert by knowledge, skill, experience, training, or education may testify in the form of an opinion or otherwise if the proponent demonstrates to the court that it is more likely than not that:

(a) the expert's scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue; (b) the testimony is based on sufficient facts or data; (c) the testimony is the product of reliable principles and methods; and (d) the expert's opinion reflects a reliable application of the principles and methods to the facts of the case.

Two changes, both deliberate:

  1. The preponderance standard is now explicit in the text. It was always the law under Fed. R. Evid. 104(a) and Bourjaily v. United States, 483 U.S. 171 (1987), but many courts had drifted toward admitting testimony and letting the jury sort it out.
  2. Subsection (d) was rewritten from "the expert has reliably applied the principles and methods to the facts of the case" to "the expert's opinion reflects a reliable application of" them. The Committee Note explains this "emphasizes 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."

The Committee Note is unusually blunt about why: "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)."

Part I: The Daubert trilogy, still governing

Daubert (1993)

Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993), held that the Federal Rules of Evidence superseded the "general acceptance" test of Frye v. United States, 293 F. 1013 (D.C. Cir. 1923), and assigned trial judges a gatekeeping role: to ensure that scientific testimony is both relevant and reliable.

The Court offered a non-exclusive list of reliability considerations:

  1. Whether the theory or technique can be and has been tested.
  2. Whether it has been subjected to peer review and publication.
  3. The known or potential rate of error.
  4. The existence and maintenance of standards controlling the technique's operation.
  5. Whether the theory or technique enjoys general acceptance in the relevant scientific community.

The Court emphasized these are "flexible" and that the inquiry focuses "solely on principles and methodology, not on the conclusions that they generate." That last sentence has been quoted more than any other in the opinion, and it was substantially qualified four years later.

Joiner (1997)

General Electric Co. v. Joiner, 522 U.S. 136 (1997), did two things.

First, it held that abuse of discretion is the standard of review for admissibility rulings, including rulings that are outcome-determinative. That gave district judges wide latitude and made exclusion rulings very hard to reverse.

Second, and more importantly for the substance, it rejected the strict separation between methodology and conclusions: "conclusions and methodology are not entirely distinct from one another. ... A court may conclude that there is simply too great an analytical gap between the data and the opinion proffered." Id. at 146.

The analytical gap language is the ancestor of the 2023 amendment, and it is the phrase to use in any motion to exclude.

Kumho Tire (1999)

Kumho Tire Co. v. Carmichael, 526 U.S. 137 (1999), extended the gatekeeping obligation to all expert testimony, not merely scientific testimony, covering technical and other specialized knowledge. It also confirmed that the Daubert factors are neither mandatory nor exclusive, and that a court has "considerable leeway in deciding in a particular case how to go about determining whether particular expert testimony is reliable."

Kumho Tire matters enormously in commercial litigation, where most experts are not scientists: accountants, appraisers, industry practitioners, software engineers, and marketing professionals. Their methods are still subject to reliability scrutiny, adapted to the field.

Weisgram (2000)

Weisgram v. Marley Co., 528 U.S. 440 (2000), held that a court of appeals that finds expert evidence was erroneously admitted may direct entry of judgment as a matter of law rather than remanding for a new trial. The practical message: a proponent who relies on inadmissible expert testimony may not get a second chance to build the case.

Part II: What the amendment changes in practice

The "weight not admissibility" reflex is now a citable error

Before December 2023, opposition briefs routinely recited that gaps in an expert's data, questionable assumptions, and disputed applications "go to weight, not admissibility." That sentence is now met with the Committee Note calling it incorrect.

For movants: cite the Committee Note by name. Frame the challenge as an application challenge under Rule 702(d), and connect it to Joiner's analytical gap.

For proponents: the response is not to repeat the old formulation but to show the application is reliable: that the expert used the right inputs, followed the methodology's own requirements, accounted for confounders and alternatives, and did not extrapolate beyond what the data support.

Overstatement is now an independent ground

The Committee Note singles out overstatement: "expert testimony may not be admitted unless the expert's opinion is limited to what can be reasonably concluded from a reliable application of the methodology." An expert who has done reliable work but states the conclusion with more certainty than the method supports may be limited rather than excluded.

This shows up frequently in forensic disciplines ("to a reasonable degree of certainty this is a match"), in causation opinions, and in damages testimony that states a precise figure derived from a range of estimates. Expect more rulings that narrow rather than exclude.

It has not become a mini-trial on the merits

The amendment did not authorize courts to choose between competing reliable experts. Where two qualified experts apply accepted methods reliably and reach different conclusions, both are admissible and the jury decides. Courts continue to say so, and proponents should say so too.

It applies to all experts, not just scientists

Damages experts, industry-practice experts, survey experts, and technical experts all face the same standard. The reliability factors are field-adapted. For a worked application in trademark surveys, see Daubert Challenges to Consumer Survey Experts in Trademark Litigation and Consumer Survey Expert Methodology in Trademark Cases.

Part III: Disclosure under Rule 26(a)(2)

Admissibility is only half the problem. Disclosure defects are the other half, and they are more often fatal.

Who must be disclosed

All witnesses who will offer opinions under Rules 702, 703, or 705 must be disclosed. Rule 26(a)(2)(A).

Retained experts: the Rule 26(a)(2)(B) report

A witness "retained or specially employed to provide expert testimony in the case," or "one whose duties as the party's employee regularly involve giving expert testimony," must provide a written report containing:

  1. A complete statement of all opinions and the basis and reasons for them.
  2. The facts or data considered in forming them.
  3. Any exhibits that will be used to summarize or support them.
  4. The witness's qualifications, including all publications in the previous 10 years.
  5. A list of all other cases in which the witness testified as an expert at trial or by deposition in the previous 4 years.
  6. A statement of compensation for the study and testimony.

"Complete statement" means complete. An opinion not in the report is generally not admissible.

Non-retained experts: the Rule 26(a)(2)(C) disclosure

For witnesses who were not retained (treating physicians, employees with relevant expertise, government investigators), the party must disclose:

  • the subject matter on which the witness is expected to present evidence under Rules 702, 703, or 705; and
  • a summary of the facts and opinions to which the witness is expected to testify.

No full report is required, and this is a meaningful cost saving. But the distinction is heavily litigated. An employee who becomes an expert for the litigation, or a physician who forms opinions on causation after being retained rather than during treatment, may cross the line into (B) territory.

Timing

Absent a stipulation or court order, disclosures are due 90 days before trial, and rebuttal disclosures within 30 days of the other party's disclosure. Rule 26(a)(2)(D). Scheduling orders almost always supersede this.

The 2010 protections

Rule 26(b)(4) protects:

  • Draft reports of any kind, regardless of form. Rule 26(b)(4)(B).
  • Communications between the party's attorney and a retained expert, regardless of form, with three exceptions: communications that (i) relate to the expert's compensation, (ii) identify facts or data the attorney provided and the expert considered, or (iii) identify assumptions the attorney provided and the expert relied on. Rule 26(b)(4)(C).

This amendment ended a decade of wasteful practice in which lawyers avoided all written communication with experts. It did not make everything privileged. The three exceptions are broad, and "facts or data considered" covers material the expert reviewed and rejected.

Consulting (non-testifying) experts are protected still further: discovery is available only on a showing of exceptional circumstances. Rule 26(b)(4)(D). Using a consulting expert to explore theories before designating a testifying expert remains sound practice.

Rule 37(c)(1): the sanction

"If a party fails to provide information or identify a witness as required by Rule 26(a) or (e), the party is not allowed to use that information or witness to supply evidence on a motion, at a hearing, or at a trial, unless the failure was substantially justified or is harmless."

Exclusion is automatic and self-executing; no motion is required, though one is usually filed. Courts weigh factors including surprise, ability to cure, disruption of trial, importance of the evidence, and the explanation for the failure.

For the full framework, see Rule 37(c)(1) Preclusion of Undisclosed Evidence.

Part IV: Rules 703, 704, 705, and 706

Rule 703 (bases). An expert may base an opinion on facts or data the expert "has been made aware of or personally observed," and if experts in the field would reasonably rely on such facts or data, they "need not be admissible for the opinion to be admitted." But inadmissible bases may be disclosed to the jury only if their probative value in helping the jury evaluate the opinion substantially outweighs their prejudicial effect, a reverse-403 balancing that favors exclusion.

This is a recurring battleground: an expert cannot be used as a conduit to put otherwise inadmissible hearsay before the jury.

Rule 704 (ultimate issue). Opinion testimony is not objectionable merely because it embraces an ultimate issue. The exception in Rule 704(b) applies only in criminal cases regarding a defendant's mental state.

But there is a real limit: an expert may not testify to legal conclusions. "The defendant breached the contract" and "this conduct violated the statute" are for the court and jury. "The defendant's practices departed from the industry standard of care in the following respects" is admissible. Draft reports with that line in mind.

Rule 705. An expert may state an opinion without first testifying to the underlying facts, but may be required to disclose them on cross-examination. Practically, this means the cross-examiner controls when the weaknesses surface.

Rule 706 (court-appointed experts). Rarely used, but available, and worth proposing in cases involving highly technical disputes where the court is likely to distrust both sides' experts.

Part V: The mechanics of a good expert practice

Selecting the expert

  • Qualifications must match the specific opinion, not the general field. A cardiologist is not automatically qualified to opine on a device's software failure mode.
  • Prior testimony is discoverable and searchable. Read the last five transcripts and every exclusion order. An expert who has been excluded before will face that history in every subsequent case.
  • Prior inconsistent publications are the most effective cross material there is.
  • Availability and communication skill matter as much as credentials. A brilliant expert who cannot explain the method to a jury is a liability.

Working with the expert

  • Do not draft the report. Even though drafts are protected, an expert who did not write their own opinion collapses on cross when asked about the analysis.
  • Identify assumptions explicitly. Assumptions counsel provided that the expert relied on are discoverable under Rule 26(b)(4)(C)(iii); pretending otherwise is worse than disclosing them.
  • Make sure the expert actually reviewed what the report says they reviewed. This is a standard deposition question and a standard place to lose credibility.
  • Test the opinion against the amended rule before the report is served: are the facts sufficient, is the method reliable, and does the opinion stay within what the method supports?

Deposing the opposing expert

Objectives, in order:

  1. Lock the opinions. "Are those all of your opinions in this case?" Everything not stated is subject to exclusion later.
  2. Establish the assumptions, and get concessions that the opinion changes if an assumption fails.
  3. Explore the method's own requirements. Every methodology has published standards; ask whether the expert followed them and, if not, why.
  4. Probe alternatives. What other explanations were considered and ruled out, and how?
  5. Find the analytical gap. Walk from data to conclusion step by step and identify where the inference is unsupported.
  6. Establish scope limits. What is the expert not offering an opinion on?
  7. Do not argue. You are building a Rule 702 motion, not winning a debate.

See The Art and Science of Depositions in Federal Civil Litigation and Preparing to Take a Deposition.

Moving to exclude

  • Timing. File with or before summary judgment where the expert's opinion is necessary to create a genuine dispute. Courts should resolve the challenge rather than assume admissibility. See Summary Judgment Under Rule 56.
  • Vehicle. A motion in limine or a separate Rule 702 motion, depending on local practice.
  • Ask for the right relief. Full exclusion is rarely granted; partial exclusion and limitation are common. Identify which opinions and which sentences.
  • Request a hearing where the reliability question turns on how the expert actually performed the analysis. Courts have discretion whether to hold one, and a well-supported request helps.
  • Support with the record. Cite deposition testimony where the expert conceded the gap, and the methodology's own published standards.

Part VI: Damages experts

Damages testimony draws more exclusion motions than any other category in commercial litigation, and the 2023 amendment has made those motions more effective.

Common defects:

  • Unsupported assumptions. A lost-profits model built on a growth rate the expert took from management projections without testing them.
  • No apportionment. Attributing all of a product's profits to one infringed feature. This is fatal in patent cases, where apportionment to the smallest salable patent-practicing unit and the entire market value rule are heavily litigated, and it is increasingly required in trademark and trade secret cases too. See Damages Apportionment in Trademark Cases.
  • Benchmark mismatch. Comparable licenses that are not comparable, or a yardstick company that differs materially.
  • Causation gaps. Correlating a revenue decline with the defendant's conduct without ruling out market conditions, competition, or the plaintiff's own decisions.
  • Double counting across damages theories.
  • Speculative new-business projections, particularly for startups without operating history.

For proponents: have the expert state the assumptions, explain why each is reasonable, test sensitivity, and show what happens if an assumption fails. A model with an explicit sensitivity analysis is far harder to exclude than one with a single number.

Part VII: Cross-examination at trial

By the time an expert reaches the stand, the exclusion motion has been decided. Cross-examination has different goals.

The four productive lines:

  1. Concession. Get the expert to agree with the propositions your expert needs. Experts in the same field agree about more than the retained positions suggest, and a series of agreements early builds credibility for the disagreements later.
  2. Assumptions. Show what the opinion depends on, and that the jury, not the expert, decides whether those facts are true. "If the jury finds X did not happen, your opinion changes, correct?"
  3. Scope. Establish what the expert did not do, did not review, and does not opine on.
  4. Bias, in moderation. Compensation and repeat-player status are fair game, but heavy-handed attacks on a credentialed witness often backfire. One clean exchange about the hourly rate and the percentage of income from litigation work is usually enough.

What rarely works: trying to out-expert the expert, arguing methodology in front of a jury, and asking open questions that invite explanation.

Checklists

Proponent's checklist

  • Qualifications match the specific opinions, not just the field.
  • Review the expert's prior transcripts, publications, and exclusion history.
  • Confirm the report contains a complete statement of every opinion.
  • List all facts and data considered, including materials reviewed and rejected.
  • Identify assumptions provided by counsel that the expert relied on.
  • Verify the 10-year publication list and 4-year testimony list.
  • State compensation.
  • Test each opinion against Rule 702(a) through (d) individually.
  • Check for overstatement: does the conclusion exceed what the method supports?
  • Confirm no legal conclusions are stated.
  • Calendar disclosure deadlines and supplementation obligations under Rule 26(e).

Challenger's checklist

  • Verify disclosure compliance; Rule 37(c)(1) exclusion is often easier than Rule 702 exclusion.
  • Identify each opinion separately; challenge them individually.
  • Map the analytical chain from data to conclusion and locate the gap.
  • Obtain the methodology's own published standards and compare.
  • Depose to lock opinions, assumptions, and scope.
  • Frame the motion under Rule 702(b) (insufficient facts) and (d) (unreliable application), citing the 2023 Committee Note.
  • Consider partial exclusion and limitation, not just wholesale exclusion.
  • Coordinate timing with summary judgment.
  • Preserve objections at trial even after a denied motion in limine.

A worked example

Ashford Materials, Inc. (fictional) sues Delta Coatings Corp. (fictional), alleging that Delta's protective coating failed and caused corrosion in Ashford's storage tanks. Ashford offers two experts.

Expert 1: a metallurgist who opines that the corrosion was caused by a defect in the coating's cure chemistry. Her report describes a standard failure analysis: visual inspection, cross-sectional microscopy, elemental analysis, and comparison to specification.

Challenge analysis. The methodology is standard and passes Rule 702(c). The fight is under (b) and (d). Delta's deposition establishes that she examined samples from two of eleven tanks, that both were from the same production lot, that she did not test the substrate preparation (a known alternative cause), and that Ashford's own maintenance records show the tanks were cleaned with a solvent the coating manufacturer's data sheet prohibits.

Under the amended rule, this is a strong motion: the opinion may not "reflect a reliable application" of failure analysis, because the method's own protocol requires ruling out alternative causes, and she did not address the prohibited solvent at all. The likely outcome is not exclusion of everything but a ruling limiting her to what her samples support, and possibly excluding the general causation opinion as to all eleven tanks.

Expert 2: a damages accountant who calculates $8.4 million in lost profits by taking Ashford's five-year revenue projection, subtracting actual revenue, and applying a 34 percent margin.

Challenge analysis. This is the classic damages exclusion candidate. He did not test the projection, did not adjust for a competitor's entry into the market during the damages period, did not account for Ashford's decision to close a distribution center, and applied a company-wide margin rather than a product-specific one. Each is an application defect under Rule 702(d), and together they are an analytical gap under Joiner.

Ashford's better approach, had it been taken at the report stage: have the metallurgist sample across lots and expressly address the solvent issue, and have the accountant build a model with stated assumptions, a sensitivity analysis, and an explicit adjustment for the competitor's entry. Both opinions would then be admissible, and the disputes would go to the jury.

The lesson. Most exclusions are earned during report preparation, not during briefing. The amended rule rewards experts who show their work and punishes those who assert conclusions.

Frequently asked questions

Did the 2023 amendment change the substantive standard? The Advisory Committee said it clarified rather than changed, because the preponderance standard always applied under Rule 104(a). In practice it has changed outcomes, because it removed the "goes to weight" shortcut that many courts had used.

Can a court exclude an expert just because it disagrees with the conclusion? No. The court assesses reliability, not correctness, and competing reliable experts both get admitted. But Joiner permits exclusion where the gap between data and conclusion is too great, and the amendment reinforces that.

Do I need an expert at all? For technical causation, professional standard of care, most damages models, and specialized industry practice, usually yes. For matters within ordinary understanding, no, and offering an expert on such matters risks exclusion under Rule 702(a) for failing to help the trier of fact.

Can my company's employee serve as an expert? Yes, and a full Rule 26(a)(2)(B) report is required only if the employee's duties "regularly involve giving expert testimony." Otherwise a Rule 26(a)(2)(C) disclosure suffices. Employee experts face credibility challenges but can be highly effective on industry practice and internal systems.

Are drafts of the expert report discoverable? No. Rule 26(b)(4)(B) protects drafts regardless of form. Communications between counsel and a retained expert are also protected, except as to compensation, facts or data provided and considered, and assumptions provided and relied upon.

What happens if I miss the expert disclosure deadline? Rule 37(c)(1) excludes the testimony unless the failure was substantially justified or harmless. Courts do grant relief, but do not count on it, especially close to trial.

Should I file my exclusion motion before or with summary judgment? With, or before. If the opposing expert's opinion is what creates the genuine dispute, the court needs to resolve admissibility to rule on summary judgment.

Can experts testify about the law? Not about legal conclusions or the governing standard, which is the court's province. They may testify about industry standards, regulatory practice, and custom. In some specialized contexts (foreign law, patent office practice), courts permit more latitude.

What is the standard of review on appeal? Abuse of discretion, under Joiner. That is a difficult standard, which is why exclusion fights are essentially decided in the district court.

Closing thought

The 2023 amendment is best understood not as a new rule but as a correction to a habit. Judges are busy, expert disputes are technical, and "the jury can sort it out" is an appealing way to move a case along. Daubert assigned a different job, and the Advisory Committee decided to say so in the text rather than in a note nobody reads.

For litigators, the operational consequences are concrete. Reports must show the work, not merely the conclusion. Assumptions must be stated rather than buried. Alternative causes must be addressed rather than ignored. Damages models need sensitivity analysis. And opinions must stop where the method stops.

Those are also the qualities that make expert testimony persuasive to a jury. The rule and the craft point the same direction, which is not always true in evidence law.

Appendix: expert practice by subject area

The reliability inquiry is field-adapted, and knowing what a court will look for in a given discipline is most of the preparation.

Technical and software experts. Courts ask whether the expert examined the actual accused system or relied on documentation, whether testing was performed and whether it was reproducible, and whether source code review was done under the protective order's procedures. An opinion that a system "must work this way" based on marketing materials is vulnerable. An opinion supported by code citations, logs, and a documented test protocol is not. See Legal Protection of Software.

Patent experts. Two distinct roles: the technical expert who addresses infringement and validity from the perspective of a person of ordinary skill in the art, and the damages expert who builds the royalty or lost-profits model. The technical expert must define the level of ordinary skill and apply the court's claim constructions rather than their own. The damages expert must apportion. Both are routinely challenged. See Comprehensive Guide to Patent Infringement Litigation.

Trademark survey experts. Reliability turns on the universe surveyed, the sampling method, the stimulus, the control, the question wording, double-blind administration, and the coding of verbatim responses. Each is an established professional standard, and a departure from any of them is an application defect under Rule 702(d).

Trade secret experts. Often two: one on whether the information is generally known or readily ascertainable, and one on damages or head-start period. The first must actually survey the public literature rather than assert unfamiliarity. See Trade Secret Misappropriation Litigation Under the Defend Trade Secrets Act.

Forensic and digital evidence experts. Chain of custody, imaging methodology, hash verification, and tool validation. Courts have grown sophisticated about the difference between running a commercial tool and interpreting its output. Overstatement is the recurring problem: "the file was deliberately deleted" is often more than the artifacts support, whereas "the file was deleted, and the following artifacts are consistent with the use of a wiping utility" is defensible.

Accounting and valuation experts. Method selection (income, market, or asset approach), the reasonableness of inputs, the treatment of risk, and consistency with the expert's own prior work. Cross-examination with the expert's published opinions or prior testimony taking a different position is devastating and common.

Industry-practice experts. These witnesses are qualified by experience under Rule 702's "knowledge, skill, experience, training, or education." Kumho Tire permits it, but the expert must explain how the experience leads to the conclusion, why it is a sufficient basis, and how it is reliably applied to the facts. An expert who says "in my thirty years I have seen this many times" without more is precisely what the amendment targets.

Medical and causation experts. General causation (can this agent cause this harm) and specific causation (did it cause this plaintiff's harm) are distinct, and an expert qualified for one may not be qualified for the other. Differential diagnosis is an accepted methodology, but it requires actually ruling in and ruling out alternatives with support.

Experts on foreign law. Fed. R. Civ. P. 44.1 permits the court to consider any relevant material in determining foreign law, and treats the determination as a ruling on a question of law. Expert declarations are the standard vehicle, and they are not subject to the same Rule 702 rigor because the audience is the judge.

Related articles

This article is provided for general informational purposes and does not constitute legal advice. Expert practice varies by court, judge, and subject matter. Consult qualified litigation counsel about any particular expert or challenge.