Vetting an Expert Witness

The rule already wrote the checklist

Whatever an expert is eventually asked to produce under FRCP 26(a)(2)(B) is what counsel should ask for at the outset, because it will be produced anyway and because the exercise of producing it early reveals how organized the candidate is. The rule requires the report to contain “the witness's qualifications, including a list of all publications authored in the previous 10 years”; “a list of all other cases in which, during the previous 4 years, the witness testified as an expert at trial or by deposition”; and “a statement of the compensation to be paid for the study and testimony in the case.”

Read those three items as a floor. The four-year window and the ten-year window are what must be disclosed; they are not what an adversary will search. A vetting exercise that stops at the rule's periods stops well short of where a competent cross-examination begins.

Vetting also has a second audience. Everything found in this process is found again, later, by the other side. The purpose is not only to decide whether to retain but to know in advance what the impeachment file will contain, so that nothing in it arrives for the first time during a deposition.

Prior testimony, and where the record actually lives

A candidate's litigation history sits in several places, none of which is complete on its own.

• Published opinions, searched by the candidate's name rather than by subject.
• Dockets — federal through PACER and commercial docket services, state through their own systems. Dockets capture cases that produced no published opinion, which is most of them.
• Full-text motion and brief databases. A motion in limine naming the candidate sets out, in an adversary's own words and at an adversary's expense, exactly what was thought to be wrong with the opinion.
• Transcript banks maintained by plaintiff-side and defense-side organizations, and commercial transcript services.
• Specialty databases that index case reports by expert name even where the published opinion does not name the expert. That capability is what ordinary case-law searching misses, and it is where a candidate's older work surfaces.

What the record answers is whether the person has been qualified before, in what kind of case, on which side, and how consistently. A candidate whose prior testimony has been exclusively for one side of a recurring dispute is not disqualified by that fact, but the pattern will be put to them and should not be a surprise to the party who retained them.

Prior exclusions, and why silence is not clearance

A search for rulings under FRE 702 naming the candidate is the highest-value search available, because an exclusion is the one adjudicated statement about an expert's method that exists anywhere. It is also the search most likely to return nothing, for a reason that has nothing to do with the candidate: most rulings on expert admissibility are made from the bench or issued as short unpublished orders that never become searchable text. A clean result establishes that the search was clean. It does not establish that the record is.

That asymmetry cuts in a specific direction. A found exclusion is strong information and should be read in full — the ground matters enormously, since an opinion excluded because the proponent failed to disclose it on time says nothing about the expert's method, while an opinion excluded for an analytical gap between the data and the conclusion says a great deal. An unfound exclusion is not information at all, and the correct response is to ask the candidate directly whether any of their opinions has ever been limited or excluded, and to note the answer.

Publications, and everything else written under the candidate's name

Read what the candidate has written, and read what has been written against it. The ABA's chapter for litigators puts the risk in one sentence: “The last thing you want to have happen is for your expert to be impeached by his own writings or previous positions.”

The search is wider than a publication list. It covers congressional and regulatory testimony, conference papers and slides, professional blog and forum posts, patents in the candidate's own name, appearances in broadcast and news databases, and video. A candidate's own patent has been used to impeach them. So has a decade-old forum post. None of that material appears on a CV, and all of it is public.

Position drift is the specific thing to look for. An expert who has taken the opposite position in print, in a different context and for a different audience, is not necessarily inconsistent — contexts differ, and science moves — but the explanation has to exist before the deposition rather than be improvised in it. A candidate who can explain the earlier position and why this case is different is stronger for having written it. A candidate who has forgotten writing it is a problem.

Credentials, verified against the issuer

Degrees are verified against a registrar. License status, and any revocation, suspension or condition, is verified against the issuing board. Board certifications are verified against the certifying body, which is also where a lapsed certification shows up. Society memberships are verified against the society, and the distinction between a fellowship awarded on peer review and a membership available on payment of a subscription is one an adversary will draw at trial. Disciplinary history is checked wherever the profession maintains it.

This step is skipped constantly because it feels insulting to the candidate. The ABA chapter's framing is that “[s]tudies suggest that falsifying credentials on a resume is not a rare occurrence among professionals.” The cost of the check is an hour. The cost of not running it is that the verification is performed for the first time by opposing counsel, in front of the jury, on the one line of the CV that is wrong.

Public records belong in the same pass: bankruptcy filings, malpractice judgments, litigation in which the candidate was a party, and employment history that may create an undisclosed alignment with a party, an insurer or a competitor.

Reading a CV against the claimed experience

Collect the candidate's CV from more than one point in time — the version attached to a prior case filing, the version on a directory listing, the version in an old conference program — and compare them line by line. Drift between versions is the cheapest red flag available, and it is invisible to anyone looking at a single current copy. The ABA chapter's example of the genre: “At some point, did the expert change his undergraduate institution from a state college to an Ivy League university?”

Then read the current CV against the opinion the case needs. Two patterns matter. The first is a credential that is real but adjacent: a distinguished record in a neighboring subfield, with nothing in the specific area the opinion will address. The second is the litigation-support-heavy CV — a document composed largely of expert engagements, with a primary practice that is thin, stale or absent.

The second pattern is a legitimate subject of both cross-examination and discovery. In Behler v. Hanlon, 199 F.R.D. 553 (D. Md. 2001), the court ordered disclosure of the percentage of the expert's gross income for each of the preceding five years attributable to expert witness services for insurers and defense counsel in personal injury cases, together with a five-year case list and a ten-year list of the insurers for whom the expert had worked. An expert whose income is concentrated in litigation work should expect the question and should be able to answer it without discomfort.

What vetting cannot settle

Three conclusions are commonly drawn from a vetting file that the file does not support.

Rate does not predict quality. No study was located correlating an expert's hourly rate with admissibility outcomes, with surviving a challenge under FRE 702, or with any measure of testimony quality. What rate correlates with is discipline — medical specialties dominate the top of every published fee table — along with market and how long the person has been marketing themselves for litigation work. A high-billing surgeon and a moderately billing metallurgist are not two points on a quality scale; they are two labor markets. The most misleading thing a buyer can do is treat a national average as a floor and read anything below it as a discount.

Testimony experience cuts both ways, and roughly symmetrically. A candidate with a long deposition history will not be rattled, will recognize a Rule 702 motion when it arrives, and will not say something reckless in the first hour. That same candidate has a stack of transcripts in circulation, an income demonstrably concentrated in litigation work, and, under Behler, may be ordered to quantify how concentrated. A candidate with no testimony history has no impeachment file and no track record either. There is no safe position on that axis, only a choice about which cross-examination is easier to defend.

The absence of an adverse ruling is not a clean record. Most rulings under Rule 702 never become searchable.

Which leaves the signal that vetting never captures, because it does not exist on paper: what the candidate declines to say. An expert who, on a first call with the facts described, says the question is outside their field, that the available data will not support an opinion to a reasonable degree of certainty, that a particular record set would have to be reviewed first, or that the answer may come back unhelpful, has said more about their reliability than the entire file. An expert who reaches the desired conclusion during the first conversation has also said something, and it is what opposing counsel will spend the deposition proving.

Frequently Asked Questions

Published 2026-08-12