Surviving Cross-Examination

Cross-examination is the principal mechanism courts use to test expert testimony, and it is where weak or overreached opinions are most exposed. This topic covers the common lines of attack on forensic evidence, how to concede genuine limits without collapsing the opinion, and how to stay within the four corners of the report.

Last updated: 24 Jun 2026

Cross-examination is the adversarial system's main tool for testing expert evidence. Counsel on the other side asks the expert questions designed to expose weaknesses in credentials, method, analysis, and opinion. An expert who has testified accurately and stayed within the bounds of the data can survive cross-examination; an expert who has overstated, who does not know the limits of the technique, or who strays outside the report is likely to be damaged. The core discipline is knowing the difference between a question that should prompt a concession and a question that should be met with a calm, firm answer.

The pattern of cross-examination is similar across common law systems. In the United States, the Daubert standard (Federal Rules of Evidence Rule 702) and its state equivalents focus judicial attention on whether the expert's methodology is scientifically sound before the expert even reaches the stand; Frye general acceptance remains the touchstone in a minority of states. In England and Wales, the Criminal Procedure Rules 2020 Part 19 and the Law Commission's guidance on expert evidence set the duties of the expert witness explicitly. In India, the Bharatiya Sakshya Adhiniyam 2023 (BSA 2023, replacing the Indian Evidence Act 1872) and the Bharatiya Nagarik Suraksha Sanhita 2023 (BNSS 2023, replacing the Code of Criminal Procedure) govern both the admissibility of expert opinion and the conduct of examination. The questions counsel asks under cross-examination in each system map onto the same underlying vulnerabilities: credentials, method, analysis, and opinion scope.

The expert's primary obligation is to the court, not to the party that retained them. This duty, codified in England and Wales (Criminal Procedure Rules 2020 r.19.2), in Australian practice (Federal Court Rules 2011 Schedule 4), and increasingly recognised in Indian judicial guidance, shapes every answer the expert gives on cross-examination. An expert who deflects a valid challenge to protect the retaining party's case is failing the court. An expert who concedes only under pressure, and then too broadly, is also failing. The goal is accurate, bounded testimony that holds up because it is true.

By the end of this topic you will be able to:

Identify the four main lines of attack used in cross-examination of forensic experts and describe how each line maps onto specific vulnerabilities in evidence.
Explain what a partial concession is, when it is appropriate, and how to frame it without undermining the core opinion.
Describe what it means to stay within the report and explain why departing from it under cross-examination is both a professional and a tactical error.
Compare how Daubert, Frye, English Part 19, and BSA 2023 rules each shape the questions counsel will ask on methodology.
Recognise the difference between a question designed to probe a genuine limit and a question designed to produce a false concession, and formulate an appropriate response to each.

Key terms

Cross-examination: Questioning of a witness by the party opposed to the party that called the witness. In common law systems, cross-examination is not limited to matters raised in direct examination and is typically broader and more adversarial in tone.
Concession: An acknowledgment by the expert witness that a particular proposition put by counsel is correct. A partial concession accepts part of a challenge while maintaining the core opinion. A full concession abandons the opinion. Knowing which is appropriate is central to surviving cross-examination.
Scope creep: The error of allowing cross-examination to push the expert into territory outside their expertise or outside the scope of their report. Scope creep typically benefits the cross-examining party, because the expert is now testifying on ground where they are less certain.
Daubert gatekeeping: The judicial function under Daubert v. Merrell Dow Pharmaceuticals (1993) and Federal Rules of Evidence Rule 702, requiring the trial judge to assess the scientific reliability and relevance of expert methodology before the expert testifies. The criteria include peer review, known error rate, testability, and general acceptance.
The report as anchor: The principle that an expert's opinion on the stand must be consistent with, and bounded by, the expert report filed before trial. The report cannot be silently expanded or contracted during testimony; any departure must be acknowledged openly.
Overreaching: The error of stating a conclusion that the underlying data does not support. Overreaching is the single most common reason experts are damaged on cross-examination, because the gap between data and conclusion is exactly what opposing counsel will locate and expose.

The four lines of attack

Experienced counsel attacking expert testimony rarely relies on a single challenge. The four standard lines of attack are credentials, methodology, analysis, and opinion. Each targets a different layer of the expert's evidence and requires a different type of response. Understanding the structure in advance removes surprise, which is one of cross-examination's most effective tools.

Line of attack	What counsel is trying to show	Typical question form	Effective response
Credentials	The witness lacks the qualifications to give this opinion	'You have never specialised in X, have you?'	Acknowledge gaps accurately; explain what qualifications you do have that bear on this analysis
Methodology	The technique is not validated, not generally accepted, or not applied correctly	'This method was not peer-reviewed until 2019. Were you applying it in 2017?'	Explain the validation status at the time; distinguish between an evolving method and an unvalidated one
Analysis	The specific work in this case was flawed: sample handling, chain of custody, instrument calibration, or interpretation	'Your laboratory's proficiency test showed a 4% error rate in this assay, correct?'	Confirm accurate data; explain what it means for this case and what it does not mean
Opinion	The conclusion overstates what the data supports, or contradicts the expert's own prior writings or testimony	'In your 2021 paper you wrote the technique cannot distinguish between X and Y. Yet today you say it can.'	Explain what has changed, or acknowledge the tension and explain how you reconcile it

Counsel often uses all four lines in sequence, beginning with credentials to establish a general sense of limitation, then moving to methodology, then analysis, then pressing the opinion. An expert who has been shaken on credentials and methodology is less likely to hold the opinion firmly. This is partly a psychological strategy. The expert's protection is knowing that each layer is independent: a credential limitation does not automatically compromise a well-supported methodology, and a methodology challenge does not automatically void a sound analysis.

How to concede without collapsing

The most common error under cross-examination is not resisting a valid concession. It is conceding too broadly: accepting that a limitation in one part of the analysis voids the entire opinion. The expert's task is to be accurate about both what is limited and what remains intact. A partial concession, clearly framed, is more persuasive than either a flat denial or a collapse of the position.

Refusing to concede a genuinely valid point is a strategic error. Judges and juries watch whether the expert acknowledges limitations. An expert who denies obvious weaknesses loses credibility on the entire opinion. The Ikarian Reefer test, formulated in National Justice Compania Naviera SA v Prudential Assurance Co Ltd (1993) and widely cited in English courts, includes the duty to disclose all relevant material information, including limitations. The same principle appears in the Criminal Practice Directions for England and Wales, in the BNSS 2023 framework for India, and in Federal Rule of Evidence 702's requirement that the expert's opinion reflects sufficient facts and a reliable methodology.

The test for whether to concede is factual, not tactical. Ask: is the proposition counsel has put to me accurate? If yes, confirm it and explain the scope. If no, say so and explain why. If you are not sure, say so. 'I would need to check the source of that figure before I could agree' is a legitimate answer. Guessing or deferring to counsel's version of a fact to avoid conflict is not.

Staying within the report

The expert report filed before trial is the anchor for testimony. In England and Wales, Criminal Procedure Rules 2020 r.19.4 requires the report to contain a summary of conclusions and the reasons for them. In US federal practice, Federal Rule of Civil Procedure 26(a)(2)(B) requires a complete statement of all opinions and the basis for each. Under BSA 2023 and BNSS 2023 in India, the expert's written report is the primary document of record, and oral testimony is expected to be consistent with it. The report cannot be silently revised during testimony.

Cross-examination creates two directions of departure from the report. The first is expansion: counsel asks whether the expert can comment on a matter not covered in the report, usually one where the expert would be on weaker ground. The correct response is to note that the topic was outside the scope of the analysis and that the expert cannot offer a reliable opinion on it without further examination. The second is contraction: counsel presses the expert to qualify or abandon conclusions the report contains. If the data supports those conclusions, they should be maintained.

The expert also needs to resist the false concession trap: a question framed to sound like the expert should agree, but which, if accepted, would misrepresent the data. A common form is the leading question that omits a qualifier: 'So you cannot tell us with certainty that this substance is cocaine?' The correct answer is not yes or no but a precise statement: 'I can tell the court that the analysis is consistent with cocaine and that the match probability for an alternative substance is less than one in ten thousand.' Precision protects the opinion.

Methodology challenges under Daubert, Frye, and English Part 19

The jurisdiction's admissibility standard shapes what questions counsel will ask about methodology. An expert preparing to testify in a Daubert jurisdiction should expect to be challenged on each of the Daubert factors: whether the theory or technique has been tested, whether it has been peer-reviewed, what the known or potential error rate is, and whether it is generally accepted in the relevant scientific community. These questions are not merely preliminary; if the expert's answers undermine any factor, the opinion may be attacked as unreliable even if it was already admitted through a pretrial hearing.

Under Frye, still applied in New York, California, Illinois, and several other US states, the challenge is narrower: is the methodology generally accepted by the relevant scientific community? This can seem easier to meet than Daubert, but a well-prepared cross-examiner can show that acceptance is more contested than it appears. The Frye question, strictly put, is about the method, not the expert's application of it; but in practice, counsel will also attack the application in order to frame the method as accepted in principle but misapplied in practice.

In England and Wales, Criminal Procedure Rules 2020 Part 19 and the accompanying Criminal Practice Directions require the expert's report to state the field of expertise, the literature and materials relied on, and the extent to which a question falls within a range of opinion. A cross-examiner in an English court will use those declarations against the expert: if the report stated that opinion on a point falls within a range, and the expert is now defending a single point within that range as though it is the only defensible position, there is a contradiction to exploit.

India's BSA 2023 does not prescribe judicial gatekeeping criteria equivalent to Daubert. Courts admit expert opinion under Section 39 BSA 2023 (formerly Section 45 of the Indian Evidence Act 1872) when the court considers an opinion on science or art to be relevant. The weight of that opinion is determined at trial, not before it, making cross-examination the primary quality-control mechanism. Methodological challenges are therefore even more consequential in Indian proceedings, because there is no pretrial filtration equivalent to a Daubert hearing. See Admissibility Standards Around the World for a fuller comparison.

Prior writings, prior testimony, and inconsistency attacks

Experienced counsel researches the expert's prior work before trial. Published papers, conference presentations, textbook chapters, and transcripts of testimony in previous cases are all potential sources of statements the expert has made that appear to contradict the current opinion. This is one of the most destabilising lines of cross-examination because it turns the expert's own authority against the opinion.

The response to an apparent inconsistency from a prior source depends on what the inconsistency actually is. Three situations arise. First, the quoted passage has been taken out of context: the expert's published statement was about a different scenario, a different method version, or a different population. The expert should explain the context clearly and specifically. Second, the position has genuinely changed since the prior publication: new research, new validation data, or revised professional consensus has moved the field. The expert should say so, name the reason for the change, and note that evolving methodology is normal science. Third, the expert actually holds a different view in this case than the prior source suggests, and has no good explanation: this is the dangerous case, and the expert must acknowledge the tension rather than pretend it does not exist.

Preparation before cross-examination includes the expert's own review of their prior publications and testimony in cases involving the same or similar methods. Counsel will have done this review. The expert who has not done it is at a disadvantage. In jurisdictions with formal expert witness disclosure (US FRCP 26, England and Wales CPR 35 or CrPR 19), the expert's prior publications are often a listed disclosure item, which means counsel has been given a roadmap.

Practical conduct under cross-examination

The practical conduct rules for cross-examination are simple to state and require discipline to follow. Answer the question asked. Do not volunteer information not sought. Do not argue with counsel. Do not adopt counsel's phrasing if it is subtly inaccurate. Direct answers to the judge or jury, not to counsel. Take time to think before answering a complex question.

The phrasing trap is particularly common. Counsel may phrase a question so that a literal yes or no answer is slightly inaccurate. The expert is entitled to rephrase: 'The accurate way to put that is...' or 'I would say it differently: the test shows X, not Y.' Courts in all jurisdictions recognise that the expert is obliged to give accurate answers, and correcting a question's framing is not evasion.

Speed is also a trap. Cross-examiners sometimes ask questions rapidly in order to prevent the witness from thinking. The expert can and should pause before answering. A pause to consider a complex technical question is not weakness; it is accuracy. Judges routinely allow experts time to reflect. If counsel is pressing for an immediate answer to a question that requires careful consideration, the expert can say: 'That requires me to consider the data carefully. May I have a moment?' This is proper conduct in any jurisdiction.

When re-examination follows, the expert's own counsel has the opportunity to clarify matters raised on cross. The expert should not attempt to use cross-examination itself to deliver the clarification, because this leads to argument with counsel and looks evasive. The place to fix a misleading impression created by cross-examination is re-examination, not the cross-examination itself. In practice, the expert can flag to their own counsel during a break that a particular answer needs context.

Worked example

Navigating a methodology challenge in a digital forensics case

A digital forensics expert faces a sustained methodology challenge over a hash-matching tool used to identify contraband images. The scenario illustrates partial concession, staying within the report, and resisting a false concession.

The expert has testified that a hash-matching algorithm identified 847 images on the defendant's device as matching known contraband images in a law enforcement database. Cross-examination begins.

Credentials challenge. Counsel asks whether the expert has a computer science degree. The expert does not. Counsel notes this for the jury and moves on. The expert should not volunteer a defence here; on re-examination, their own counsel can elicit the relevant experience and training. Engaging in a credentials debate during cross-examination rarely helps.
Methodology challenge. Counsel puts to the expert that the hash database version used in the analysis was 14 months old at the time of the examination. The expert confirms this is accurate. Counsel asks whether newer hashes would have produced a different result. The expert's accurate answer is: 'A newer database would potentially identify additional images not identified in this examination. It would not alter the matches found, because hash values for specific images do not change. My report identifies the images matched at the time of the examination using the database version available.'
Analysis challenge. Counsel produces a technical paper reporting that hash-matching tools have a false-positive rate of 0.003% on large datasets. The expert is asked whether 847 images from a large dataset could include false positives. The partial concession is appropriate: 'That error rate is recognised in the literature. Applied to a dataset of this size, it is statistically possible that a small number of matches are false positives. That is why each flagged image was also visually reviewed by a trained analyst, as stated in my report. The final count of 847 reflects only images confirmed by visual review.' This concedes the statistical point while anchoring to the actual methodology applied.
Scope expansion attempt. Counsel asks whether the expert can tell the court who placed the images on the device. This is outside the scope of the report, which addressed only the presence of the images. The expert states: 'Attribution of who placed files on a device was not within the scope of my examination and I have not formed an opinion on that question.' This is the correct answer. Attempting to answer would be scope creep into territory the expert has not analysed.
False concession attempt. Counsel asks: 'So you cannot tell this court with certainty that these are what the prosecution says they are?' The question is designed to elicit a broad concession of uncertainty. The accurate answer is: 'I can tell the court that each of the 847 images produces a hash value that matches the hash value of a known image in the law enforcement reference database, and that each image was independently confirmed by visual review. The confidence in those matches is very high. There is always a residual uncertainty in any forensic analysis; I have described the error rate in my report.'

Check your understanding

Question 1 of 4· 0 answered

Counsel puts to a DNA expert: 'The reference database for this population group contains only 1,200 profiles, which is far smaller than recommended.' The expert agrees this is accurate. What is the appropriate next step?

Key Takeaways

Cross-examination attacks expert testimony on four layers: credentials, methodology, analysis, and opinion scope. Each layer is independent, and a weakness in one does not automatically undermine the others.
A partial concession, which acknowledges a specific limitation while maintaining the core opinion, is both more credible and more accurate than either flat denial or wholesale retreat.
The expert report is the anchor for testimony. Departing from it by expanding scope or silently contracting a conclusion during cross-examination is a professional error in every jurisdiction, whether under FRCP 26, CrPR Part 19, or BSA 2023.
Daubert and Frye jurisdictions in the US shape methodology questions toward scientific reliability criteria; England and Wales require the expert to declare ranges of opinion; India's BSA 2023 makes cross-examination the primary quality-control mechanism because there is no pretrial admissibility gatekeeping equivalent to Daubert.
Practical discipline under cross-examination means answering precisely, correcting inaccurate phrasing without argument, refusing to be hurried into imprecise answers, and leaving clarification of misleading impressions to re-examination rather than attempting to argue during cross.

What are the most common lines of attack on a forensic expert during cross-examination?

Opposing counsel typically challenges the expert on four fronts: qualifications (attacking credentials or experience), methodology (questioning whether the technique is validated or accepted), the specific analysis (pressing on sample handling, chain of custody, or interpretation choices), and the opinion itself (pushing the expert beyond what the data supports or into territory the report did not cover).

How should an expert concede a genuine limitation without undermining the whole opinion?

The expert should concede the specific limitation plainly and then state what the limitation does and does not affect. For example: 'You are correct that the reference database for this population is limited. That affects the precision of the match statistic but does not change my conclusion that the profiles are consistent.' Partial concession paired with a clear statement of what remains valid is more credible than either refusing to concede or collapsing entirely.

What does it mean for an expert to stay within the report during cross-examination?

Staying within the report means not adopting new conclusions, not expanding the scope of the opinion, and not retreating from conclusions that the data supports, all in response to counsel's questions. The report defines the expert's position. If cross-examination surfaces something genuinely new, the proper course is to say so openly rather than to silently adjust the opinion on the stand.

How do the Daubert and Frye standards shape cross-examination of forensic experts in US courts?

Under Daubert (federal courts and most US states), opposing counsel can challenge whether the expert's methodology meets the judicial gatekeeping criteria: peer review, known error rate, general acceptance, and testability. Under Frye (still used in some states), the challenge focuses narrowly on general acceptance. An expert who knows these standards can anticipate methodological challenges and answer them precisely rather than being caught off-guard.

What does the Bharatiya Sakshya Adhiniyam 2023 say about cross-examination of experts in India?

The Bharatiya Sakshya Adhiniyam 2023, which replaced the Indian Evidence Act 1872, preserves the rule that an expert's opinion is relevant but not conclusive. Section 39 (formerly Section 45) allows courts to call experts; Section 143 (formerly Section 138) permits cross-examination of any witness, including experts. Indian courts have consistently held that the weight given to expert evidence depends on how well it withstands cross-examination on the method, the basis, and the reasoning chain.

Test yourself on Forensic Law with free, timed mocks.

Practice Forensic Law questions

Found this useful? Pass it along.

Spotted an error in this page? Report a correction or read our editorial standards.

Key Takeaways

Your journey to becoming a forensic professional starts here.