Expert testimony rarely falls apart because of one dramatic admission. More often, the problem is drift: an opinion that shifts between the report, deposition, and trial, or credentials and methods that look firmer on paper than they do under scrutiny.
For litigators, the practical question is not just whether an expert changed something. It is whether the change reflects legitimate opinion development or a credibility problem worth pressing. These eight clues can help you spot the difference.
1. The opinion got stronger over time
A common red flag is escalation. The report uses qualified language, the deposition adds confidence, and trial testimony becomes absolute.
Look for:
- “Possible” becoming “probable”
- “Consistent with” becoming “caused by”
- A limited opinion expanding into broader conclusions
What it may suggest:
- Advocacy rather than independence
- Pressure from the retaining side
- An effort to fit trial themes more tightly than the underlying analysis supports
What to compare:
- Expert report
- Deposition transcript
- Trial outlines or disclosures
- Errata sheets
2. Key assumptions changed without a clear reason
Experts are allowed to revise opinions when new facts emerge. The problem is when assumptions change and the record does not show why.
Watch for shifts in:
- Exposure levels
- Timing assumptions
- Wage or damages inputs
- Alternative cause analysis
- Reliability of source data
A useful test is simple: what new information caused the change, and when did the expert receive it? If the answer is vague, the revision may be more vulnerable than it first appears.
3. The math moved, but the method supposedly did not
Sometimes the expert insists the methodology is unchanged, yet the calculations tell a different story.
That often appears in:
- Revised damages models
- New spreadsheets
- Different discount rates
- Changed sample sizes
- Reworked benchmarks
What to pull:
- All report versions
- Backup calculations
- Spreadsheets and formulas
- Data extracts
- Notes reflecting revisions
If the output changed materially, pin down whether the inputs, assumptions, or formula changed too. “Same method” is not much of an answer if the engine underneath it was rebuilt.
4. The expert now relies on materials they did not previously mention
Late-added reliance materials can signal backfilling. An expert may cite literature, records, or testing at trial that never appeared in the report or deposition.
That matters because it raises practical questions:
- When did the expert review those materials?
- Why were they omitted earlier?
- Did the materials actually inform the opinion, or are they being used to defend it after the fact?
Build a timeline of materials reviewed, opinions formed, and disclosures made. Inconsistency often becomes clearer when dates are placed side by side.
5. Prior testimony in other cases does not match the current position
An expert may take one position in one case and a different one in another, especially on methodology, causation standards, or industry practice.
Useful sources include:
- Prior deposition transcripts
- Trial testimony
- Affidavits
- Published articles
- Presentations
- Treatise authorship
Not every difference is impeachment gold. Case facts vary. But when the expert describes the same method or principle differently across cases, that can be a powerful credibility point.
6. The expert stepped outside the stated field of expertise
Scope creep often shows up gradually. A physician offers engineering opinions. An engineer speculates about future medical care. An economist starts interpreting clinical causation.
Compare:
- CV
- Board certifications
- Prior testimony history
- The exact opinions offered
The inconsistency here is between claimed expertise and actual testimony. That gap can be more damaging than a contradiction in wording because it goes to admissibility and weight at the same time.
7. The recordkeeping is thin or oddly timed
Missing notes, late-created summaries, and unexplained revisions can all support an inconsistency narrative.
Pay attention to:
- Workpapers created only after deposition
- Sparse notes for a major opinion
- Drafts with significant unexplained changes
- Incomplete testing documentation
- Invoices that do not line up with the claimed work sequence
Documentation problems do not prove the opinion is wrong. But they can make it much harder for the expert to explain when an opinion formed and why it changed.
8. The expert becomes evasive when pinned to prior statements
Often, the inconsistency is not just in the documents. It is in how the expert reacts when confronted with them.
Typical signs include:
- Repeated “I don’t recall”
- Nonresponsive answers
- Distinctions without a real difference
- Argumentative pushback on simple comparisons
That is why deposition pin-down questions matter. Lock in the opinion, the assumptions, the materials reviewed, and whether anything remains tentative. If the expert later moves, the contrast is cleaner.
Separate revision from contradiction
Not every change is impeachable. A revised opinion may be entirely fair if:
- New records were produced
- Testing was completed later
- Assumptions were expressly provisional
- The expert disclosed the update clearly and promptly
The stronger impeachment point is usually not that the opinion evolved, but that it evolved without a documented, defensible reason.
Why this matters in practice
Inconsistent expert testimony is rarely just a cross-examination issue. It can affect expert selection, deposition planning, motion practice, and trial presentation.
The most effective approach is comparative and chronological: line up the report, transcript, workpapers, prior testimony, publications, and supporting data, then ask what changed, when, and why. That is also where targeted expert witness research can help. Services like Expert Radar can assist counsel in surfacing prior testimony patterns, background issues, and other credibility signals before those problems emerge in the courtroom.
The goal is not to manufacture inconsistency. It is to identify whether the expert’s opinion is truly stable, or only looks that way from a distance.
