I have spent 29 years as a civil and structural engineer, mostly on large industrial and energy projects. This year I wrote my first expert witness report for a construction dispute. Engineering experience prepares you well for the technical work. It does not fully prepare you for writing for a court. These are the lessons I took from it.
1. Your duty is to the court
Every expert training course says this. It feels different when you are drafting. One side instructs and pays you, you work closely with their solicitors, and it is natural to feel some loyalty to them. That feeling is the risk.
I have seen several court cases where the expert was knowledgeable and experienced in their field, yet presented evidence that clearly favoured the instructing party. The technical content was often sound. The criticism was usually about selectivity: the court found that the expert had not considered all the evidence, only the material that strengthened the instructing party's position. Once a court sees that, the expert tends to be criticised and given less weight, including on the points where they were right. Experience counts for little if the court stops trusting your independence.
Staying objective and impartial takes discipline and mindfulness. It is not something you decide once at the start of an instruction. You have to keep checking yourself as the work goes on. That means being willing to revisit your position when new evidence arrives, being self-critical about your own draft, and reading it as the other side's expert would. If your view changes, you say so.
In practice this comes down to small choices: considering all the evidence, including documents that cut against the party instructing you, and explaining why you give each the weight you do; conceding what the evidence requires; avoiding the language of advocacy; and writing so that a reader cannot tell which side instructed you. A useful test is whether the report would read the same if the other side had instructed you. The most useful thing an expert can give an instructing party is credible evidence, and credibility comes from independence.
Justice is blind, and so should an expert witness be.
2. Say plainly where your expertise ends
Construction disputes cross disciplines. A single claim can touch structure, plumbing, electrics, acoustics and fire protection. I state where my expertise stops, give only what I can support from documents and observation, and recommend specialist evidence where it is needed. A clear boundary serves the court better than a confident guess, and it holds up much better under cross-examination.
That does not mean you should stay silent on anything outside your field. You can still state an opinion, as long as you make it clear that it is not your area of speciality and explain what it rests on.
Being outside my speciality did not stop me from researching and developing my knowledge. I read up on areas such as electrical installations, plumbing and the causes of tile cracking on unstable substrates. I borrowed the British Standards for tiling so that I could find the clauses describing the symptoms of an unstable or moving substrate, and my reasoning could rest on a recognised source rather than on my own impression. Research like this lets you reason sensibly about the evidence. It does not make you a specialist, and where a definitive opinion is needed, I say so.
Courts often see this openness as a strength, not a weakness. An expert who is honest about the limits of their expertise comes across as a professional, and the court tends to trust them more, even when they have also offered some views outside their claimed field.
3. Check every claim against the document it cites
Witness statements point to exhibits by page number. Read the exhibit. Dates, authors and contents do not always match the way a document is described. This is not about catching anyone out. It is about making sure the report rests on what the evidence actually says. The same rule applies to my own sources: I cite only what I have read and verified myself.
4. Separate fact, opinion and assumption
There are always gaps in the evidence. Where I have to assume something, such as the likely sequence of events between the parties, I label it as an assumption and say why. A reader should always be able to tell what I observed, what I was told and what I infer.
This is how I keep the four apart in a report:
- Observed: what I saw or measured myself on site, such as a dimension, a reading or a condition. I record it as a finding, not as an interpretation.
- Told: what a party or witness says happened. I attribute it to them and say I cannot verify it. Where accounts conflict, I do not choose between them without documents that support one.
- Opinion: an inference from what I observed, worded as an inference. I write that something is consistent with a particular cause, not that it was definitely caused by it.
- Assumption: something I have to take as true because the evidence is incomplete. I name it, say why I made it and say which conclusions depend on it.
If an assumption turns out to be wrong, the reader can see which conclusions are affected, and I can revise them.
5. Record what you did and did not see
Site inspections are limited by access, finishes and conditions on the day. I note what I could not inspect, I keep my method simple and repeatable, and I label any quick check as informal rather than letting it look like a formal test.
6. Leave legal conclusions to the court
Terms such as negligence are legal conclusions. My role is to give the engineering view on what reasonable skill and care would have looked like in the circumstances, and on what caused what. Seeing the term used in the other side's evidence was a reminder of the boundaries of my role: the engineering is mine to explain, and the legal conclusion belongs to the court.
7. Weigh other evidence by its quality
When I read a third-party inspection report, I ask when the inspection took place, what the inspector could actually see, whether there are measurements or photographs behind the conclusions, and whether the author has any relationship with a party. A report with no supporting data deserves careful handling, whoever commissioned it.
I do not accept evidence at face value. In one of the claims, an inspection report provided in the disclosures had been written by someone with a long-standing working relationship with the party relying on it, who had also supplied materials for the same project. The report was undated, did not explain the conditions under which the inspection was carried out, and included no measurements or photographs to support its conclusions. The inspection itself took place shortly after the start of the dispute. Together these made me cautious. I saw a potential conflict of interest, and I flagged that, along with my concerns about the report's reliability, in my own report.
Raising a concern like this does not mean dismissing the evidence. It means telling the court what it should bear in mind when deciding how much weight to give it.
8. Answer questions from the other side with care
Questions from the other side need careful handling. A question can pull you away from the context of the report. It may paraphrase a conclusion loosely, change a word that matters, or rest on a premise the report does not support. So I read each question slowly and go back to the passage it refers to. If a question misstates my position, I say so politely and restate it before answering. I answer in the first person, answer only what is asked and point back to the report.
Being careful does not mean being evasive. If a question shows I got something wrong, I say so, because my duty is to the court. Every conclusion should survive two questions: how do you know, and what if you are wrong?
This article reflects my experience of writing my first expert witness report. It is general in nature and does not disclose confidential information about any current or past instruction.