How to Write an Expert Witness Report That Survives a Daubert Challenge

Published September 19, 2026 · 14 min read

Most Daubert motions aren't won or lost on your credentials. They're won or lost on whether your report shows its work — whether a reader can trace a straight line from the facts you reviewed, through the method you applied, to the opinion you reached. Courts have a name for a report that skips that chain and just states a conclusion: ipse dixit, and it is the single most common reason a technically qualified expert gets excluded.

This guide covers what Rule 26(a)(2)(B) requires your report to contain, the Rule 702 reliability standard it has to clear once it's challenged, the ipse dixit problem and how to avoid it, a section-by-section structure for a defensible report, and the filing deadlines and supplementation duties that apply once it's submitted.

What Rule 26(a)(2)(B) Requires

If you're retained or specially employed to provide expert testimony (as opposed to a treating physician or other non-retained expert covered by the lighter-weight Rule 26(a)(2)(C) disclosure), your written report must contain six things:

• A complete statement of all opinions you will express and the basis and reasons for them. Not a summary — every opinion, with the reasoning attached.
• The facts or data you considered in forming those opinions.
• Any exhibits you will use to summarize or support your opinions.
• Your qualifications, including a list of all publications you authored in the preceding 10 years.
• A list of all cases in which you testified as an expert at trial or by deposition during the preceding four years.
• A statement of your compensation for the study and testimony in the case.

The last two items — the four-year testimony list and the compensation statement — are their own recurring headache the longer you practice, and we cover the mechanics of tracking them in detail in our Rule 26 disclosure guide. This guide focuses on the part of the report that actually gets challenged: the opinions and the reasoning behind them.

The Rule 702 Bar Your Report Has to Clear

Rule 26(a)(2)(B) tells you what your report has to contain. Federal Rule of Evidence 702 tells you the standard the opinions inside it have to meet once opposing counsel moves to exclude them. As amended effective December 1, 2023, Rule 702 requires the party offering expert testimony to demonstrate, by a preponderance of the evidence, that:

• Your knowledge will help the jury understand the evidence or decide a fact in issue;
• Your testimony is based on sufficient facts or data;
• Your testimony is the product of reliable principles and methods; and
• Your opinion reflects a reliable application of those principles and methods to the facts of the case.

The 2023 amendment mattered because before it, some courts treated the last two requirements as questions of weight for the jury rather than admissibility for the judge to decide — effectively letting shaky methodology in and leaving it to cross-examination to sort out. The amendment made explicit what the rule always meant: the judge has to find reliability by a preponderance of the evidence before the jury ever hears the opinion. Practically, that means your report is the primary evidence a judge uses to make that finding — not your CV, not your testimony at a later hearing.

Daubert, Kumho Tire, and the Reliability Factors

Rule 702's reliability requirement traces back to Daubert v. Merrell Dow Pharmaceuticals (1993), which gave trial judges a gatekeeping role and a non-exhaustive set of factors for assessing whether a methodology is reliable:

• Whether the method can be and has been tested;
• Whether it has been subjected to peer review and publication;
• Its known or potential rate of error, and the existence of standards controlling its operation; and
• Whether it enjoys general acceptance within the relevant field.

The Supreme Court was explicit that this isn't a checklist every report must satisfy item by item — it's a flexible set of considerations, and which ones apply depends on the field. That flexibility matters because six years later, in Kumho Tire Co. v. Carmichael (1999), the Court held that Daubert's gatekeeping duty applies to all expert testimony, not just testimony from bench-scientists. Rule 702 doesn't distinguish scientific knowledge from technical or other specialized knowledge, so an accident reconstructionist, a forensic accountant, or a nurse expert faces the same reliability gate a toxicologist does — the specific factors a judge leans on will just look different across those fields.

The Ipse Dixit Problem

Ipse dixit is Latin for “he said it himself,” and it's the term courts use for an opinion that asks the reader to accept a conclusion on the expert's authority alone. In General Electric Co. v. Joiner (1997), the Supreme Court held that a trial court may exclude expert testimony where there is “simply too great an analytical gap between the data and the opinion proffered” — meaning a judge doesn't have to accept an expert's conclusion just because the underlying facts or data are themselves legitimate. Joiner also settled a separate procedural point: appellate courts review a trial judge's decision to admit or exclude expert testimony for abuse of discretion, which is a deferential standard that makes getting the report right the first time considerably more important than hoping to win the issue on appeal.

In practice, the analytical gap shows up in reports as sentences like “based on my review of the file and my experience, it is my opinion that…” with nothing connecting the specific facts reviewed to the specific method applied to reach that specific conclusion. The fix isn't longer opinions — it's opinions that are explicit about the chain: this fact, run through this method, produces this conclusion, and here is why the method is reliable for this kind of question.

The “Reasonable Degree of Certainty” Trap

Many experts were taught to close every opinion with some version of “to a reasonable degree of [medical/scientific/engineering] certainty,” on the theory that the phrase is what makes an opinion admissible. No federal rule requires that language, and it has fallen out of favor. The National Commission on Forensic Science has taken the position that legal professionals shouldn't condition admissibility on that phrasing, partly because practitioners don't use it outside a courtroom — it's legal boilerplate grafted onto a scientific or clinical opinion, not a standard the field itself recognizes. At least one federal court, in United States v. Glynn, went further and barred the phrase outright in a firearms-identification case.

The practical takeaway: don't treat the certainty phrase as a substitute for showing your work. A report that shows a reliable method reliably applied to sufficient facts satisfies Rule 702 whether or not it uses that exact sentence. A report that skips the reasoning and leans on the phrase instead is exactly the kind of ipse dixit problem courts have gotten more comfortable calling out.

Anatomy of a Defensible Report

There's no single mandated format, but a structure that tracks Rule 26(a)(2)(B) and makes the Rule 702 reasoning chain easy for a judge to follow looks like this:

• Assignment and scope. Who retained you, what question you were asked to answer, and what you were not asked to address — scoping the assignment up front heads off a later argument that you strayed outside your expertise.
• Qualifications. A summary tied to the specific question in this case, with your full CV and 10-year publication list attached as an exhibit rather than recited in the body.
• Materials and facts considered. An itemized list of every document, data set, deposition transcript, and site inspection you relied on — specific enough that opposing counsel can request the same materials and follow your path.
• Methodology. The method or standard you applied and why it's reliable for this type of question — this is the section a Daubert motion will target first, so it deserves more than one sentence.
• Opinions. Each opinion stated separately, with an explicit link back to the specific facts and the specific step in your methodology that produced it — this is where ipse dixit gets fixed or doesn't.
• Exhibits. Numbered and referenced by number in the opinions section above, not described loosely. If your case reaches deposition, these report exhibit numbers and the case's own exhibit numbers are two different systems — see our guide to organizing deposition exhibits for how to keep the two straight.
• Compensation and prior testimony. Your rate structure and the four-year trial/deposition list required by Rule 26(a)(2)(B) — the mechanics of maintaining that list are covered in full in our Rule 26 disclosure guide.
• Signature and certification. Most jurisdictions and standing orders expect the report to be signed by the expert personally, not just prepared by them.

Filing Deadlines and the Duty to Supplement

Rule 26(a)(2)(D) sets a default deadline of at least 90 days before trial (or the trial-ready date) if the court hasn't set its own schedule, with rebuttal reports due within 30 days after the opposing party's disclosure. In practice almost every federal case has a scheduling order that sets its own expert-disclosure dates, and that order controls over the default — check it before assuming the 90-day rule applies.

The obligation doesn't end at submission. Rule 26(e) requires you to supplement your report if you later learn the disclosure was incomplete or incorrect, or if your opinion changes based on new information. Under Rule 37(c)(1), failing to disclose or supplement as required generally means you can't use that information or witness at trial, unless the failure was substantially justified or harmless — the same exclusion risk that applies to a late or incomplete report in the first place.

Pre-Submission Checklist

• Does every opinion trace explicitly back to a specific fact and a specific step in your stated methodology — no unexplained jumps?
• Is the methodology section long enough to explain why the method is reliable for this question, not just what the method is?
• Are all materials considered itemized, not summarized as “case file”?
• Are exhibits numbered and cross-referenced by number inside the opinions section?
• Is the four-year testimony list current as of the disclosure date, not reconstructed from memory?
• Have you checked the case's scheduling order for its actual disclosure deadline, rather than assuming the 90-day default applies?

Before your report gets tested at deposition, it's worth a final read as if you were opposing counsel looking for the weakest sentence — the same first step we recommend in our deposition preparation guide. A report that survives that read on your own desk is a report that's already most of the way to surviving a Daubert motion. And your report isn't the only document a Daubert challenge can target — if opposing counsel goes after your fees or your billing records instead of your methodology, that's a separate fight, covered in our guide to defending expert witness fees against Daubert challenges.

Frequently Asked Questions

What must an expert witness report include under FRCP 26(a)(2)(B)?

A complete statement of all opinions the expert will express and the basis and reasons for them; the facts or data considered in forming those opinions; any exhibits used to summarize or support them; the expert's qualifications, including a 10-year publication list; a list of all cases in which the expert testified as an expert at trial or by deposition in the preceding four years; and a statement of the expert's compensation for the study and testimony.

What is “ipse dixit” and why does it get expert testimony excluded?

Ipse dixit — Latin for “he said it himself” — describes an opinion offered on the expert's own authority without a shown chain of reasoning connecting the methodology and facts to the conclusion. In General Electric Co. v. Joiner (1997), the Supreme Court held that a trial court may exclude expert testimony where there is simply too great an analytical gap between the data and the opinion offered — the report asserts a conclusion rather than demonstrating it.

Does an expert report need to state opinions “to a reasonable degree of certainty”?

No federal rule requires that phrasing, and it has drawn increasing criticism. The National Commission on Forensic Science has taken the position that experts shouldn't be required to invoke a “reasonable degree of scientific certainty,” since practitioners don't use that language outside a courtroom, and at least one federal court (United States v. Glynn) barred the phrase outright in a firearms-identification case. The 2023 amendment to Rule 702 puts the emphasis where it belongs: on showing a reliable application of your methodology to the facts, not on invoking a certainty phrase as a substitute for that showing.

When is an expert report due before trial?

Rule 26(a)(2)(D) sets a default of at least 90 days before the trial date or the date the case is set for trial, with rebuttal reports due within 30 days after the other party's disclosure. In practice, nearly every case has a scheduling order that sets its own expert-disclosure deadlines, and the scheduling order controls over the default.

What happens if an expert report is late or incomplete?

Under Rule 37(c)(1), a party that fails to disclose information required by Rule 26(a) — including a complete, timely expert report — generally cannot use that witness or information at trial unless the failure was substantially justified or harmless. There's also an independent duty to supplement a report under Rule 26(e) if the expert later learns new information or changes an opinion; an unsupplemented report faces the same exclusion risk.

Key Takeaways

• Rule 26(a)(2)(B) requires six things in a retained expert's report: opinions with basis and reasons, facts/data considered, exhibits, qualifications with a 10-year publication list, a four-year testimony list, and a compensation statement.
• Rule 702 (as amended 2023) requires the proponent to show, by a preponderance of the evidence, sufficient facts/data, a reliable method, and a reliable application of that method to the facts — not just that the method exists.
• Daubert's reliability factors are flexible and non-exhaustive; Kumho Tire (1999) confirmed they apply to all expert testimony, not just hard science.
• “Ipse dixit” — an unexplained analytical gap between data and conclusion — is the single most common reason a qualified expert's opinion gets excluded (General Electric Co. v. Joiner).
• No rule requires “to a reasonable degree of certainty” language, and some courts have rejected it outright; showing reliable methodology matters more than the phrase.
• Check the case's scheduling order for the real disclosure deadline — the 90-day default in Rule 26(a)(2)(D) only applies when the court hasn't set one.
• The duty to supplement under Rule 26(e) doesn't end at submission, and Rule 37(c)(1) can exclude testimony over a violation that wasn't substantially justified or harmless.