One distinction decides almost everything about how a genetic test sits on a UK life insurance application: was it predictive or diagnostic? A predictive genetic test looks for an inherited fault in someone who has no symptoms, to estimate future risk. A diagnostic genetic test is done to explain symptoms you already have, or to confirm a condition. The first is protected and cannot be asked for; the second is ordinary medical evidence and is disclosable if the insurer asks. Get that one classification right and most of the confusion falls away.

The trouble is that the label is not written on your results letter, and the review of the UK Code found that most people closely involved in this subject cannot reliably apply the definitions to real cases. So this page does the sorting for you. Below are ten real situations, each pushed onto one side of the line or the other, with the reason spelled out — because the reason is what lets you classify the eleventh case you meet yourself.

At a glance

  • Predictive test, no symptoms, protected: the insurer cannot ask for or use the result. Diagnostic test to explain symptoms, disclosable: it is medical history.
  • The deciding question is why was the test done and did you have symptoms at the time — not who carried it out.
  • NHS, private clinic and shop-bought consumer kits are treated the same way. The route does not change the answer.
  • A cascade test after a relative's diagnosis is predictive in almost every case, so it is protected.
  • The one predictive result an insurer may ever ask about is Huntington's disease, and only above £500,000 of life cover.

The line everything turns on

Underwriting a life policy is an exercise in judging risk, and the Code on Genetic Testing and Insurance draws a firm boundary around one type of information. A predictive genetic test — carried out on a healthy person to find an inherited variant linked to a future condition — is the thing insurers have agreed not to ask for and not to use. That protection exists so that people are not deterred from tests that could help them, by the fear that a result would count against them for cover.

A diagnostic test is outside that protection because it is part of ordinary medicine. If you already have symptoms, and a genetic test is used to name or confirm what is causing them, the result is simply part of your diagnosis, and diagnoses are disclosable when an insurer asks about your health. The genetic technology is the same in both cases; the classification turns entirely on the clinical purpose and on whether you were unwell when it was done.

Because that boundary carries so much weight, it helps to be precise about what a test was actually for before you answer anything, since the language around genetics is used loosely in everyday speech and a single word on a form can change the classification. It also helps to know that this boundary is fairly new: the current Code replaced an older voluntary arrangement, and the shift is worth understanding, which we cover in what replaced the genetic testing moratorium in 2018.

Five results that stay off the form: predictive

1. A cascade BRCA test after your mother's diagnosis, with no symptoms of your own

This is the textbook predictive test. Your mother's cancer revealed a known BRCA fault, the clinic tested you to see whether you inherited it, and you were well at the time. Whatever the result, it is a predictive genetic test and cannot be asked for or used on a UK life application. It stays off the form.

2. A predictive Huntington's disease test in an affected family, while well

Also predictive, and protected in the same way — with one narrow qualification that applies to no other condition. Huntington's disease is the single result an insurer may ask about, and only once the total life cover you are applying for exceeds £500,000. Below that figure it is protected like any other predictive result, and above it, it is the only genetic result that becomes askable at all.

3. A Lynch syndrome test prompted by the family, with no bowel symptoms

Predictive. You are being tested because the condition runs in the family, not because you are unwell, so the result is protected. Note the separation clearly: the protected thing is the gene result. If that result later leads to regular colonoscopies, the fact that you are under surveillance is ordinary medical history and can be asked about, even though the gene test behind it cannot.

4. A 23andMe or Ancestry report flagging a variant, with no symptoms

Predictive, and treated exactly like an NHS or private clinical predictive test. The Code makes no distinction based on who did the test, so a consumer kit result carries the same protection as one from a hospital genetics service. You need not volunteer it, and an insurer will not ask for it. Where the report has frightened you, it is worth knowing the science rarely justifies the fright, a point we return to below.

5. A carrier test for a recessive condition, done for family planning

Generally protected, and often simply not relevant. A carrier test tells you whether you could pass a recessive condition — cystic fibrosis, say — to a child if your partner is also a carrier. It usually says nothing about your own future health, so it is not a predictive test for a disease you will develop, and it is not something insurers ask about for your own mortality.

Five results that go on it: diagnostic and disclosable

6. Genetic profiling of a tumour after a cancer diagnosis

Diagnostic, and disclosable. If you have already been diagnosed with cancer and the tumour is genetically profiled to guide treatment, that is part of your diagnosis and care, not a predictive test on a healthy person. The cancer itself is plainly disclosable, and the genetic analysis of it travels with the diagnosis.

7. A gene test that confirms a cardiomyopathy you have symptoms of

Diagnostic. If you were breathless or had an abnormal heart scan, and a genetic test was used to confirm an inherited cardiomyopathy behind those symptoms, the result explains an existing condition. It is medical evidence and is disclosable when the insurer asks about your heart. Contrast this with a well relative screened only because of a family death — that screening is a different situation, which we walk through in the sibling case of an unexplained pattern across three generations.

8. A test to explain a child's or your own developmental condition

Diagnostic. Where a genetic test is used to identify the cause of a condition that is already present and causing problems, it is confirming a diagnosis rather than predicting a future one. The condition and its investigations are disclosable medical history.

9. Surveillance or surgery that followed a predictive result

A split case worth stating carefully. Suppose a predictive BRCA result led to annual screening or to risk-reducing surgery. The gene result stays protected and cannot be asked for. But the screening programme, the operations and the hospital admissions are ordinary medical history, and they must be disclosed if the insurer asks about treatment or operations. Answer factually about the surgery; keep the test result itself out of it.

10. A test done to explain symptoms that came back negative

The gene result may be negative and protected, but do not mistake that for having nothing to declare. If you had symptoms investigated — chest pain, a collapse, unexplained bleeding — those symptoms and investigations are medical history in their own right, regardless of the genetic result. What gets disclosed is the clinical picture, not the gene.

The edges the ten examples expose

Three patterns run through the list. The first is that the route never decides it: NHS, private and consumer tests are all treated identically, so "it was only a home kit" changes nothing, and neither does "it was done privately". The second is that a predictive result almost never travels alone — it tends to generate screening, medication or surgery, and those clinical consequences are disclosable even when the result behind them is not. The third is the research carve-out, which sits outside this predictive-versus-diagnostic split entirely: a result generated in a research study is not disclosable at all, though anything clinical that follows from it is. The line between a result you must not be asked for and a decision an insurer got wrong is also its own subject, covered in what happens when a genetic test comes after cover has already started.

Underneath all of this sits a single hard threshold, which is worth stating with its source rather than from memory.

Source: Code on Genetic Testing and Insurance, HM Government and the Association of British Insurers, published on gov.uk, with its three-year review published in 2025.

What is still disclosable, whatever the test was

None of this touches the questions insurers are entitled to ask directly. Family history is one of them: an application can ask how many close relatives had a condition and at what ages, at any level of cover, and that is not a genetic test and not protected by the Code. Your own symptoms, diagnoses, medication and treatment are disclosable when asked. The protection is specific — it stops a predictive result being extracted from you or your notes — not a blanket shield over your whole medical past.

Where the two things can blur is in your GP records, because a predictive result and the clinical follow-up it triggered often sit side by side in the same notes. When an insurer asks your doctor for a report, you are entitled to see it before it goes and to have a protected predictive result removed from it, while the disclosable medical history stays. That right is more useful than most people realise, and we set it out step by step in how to read the GP report before it is sent. Used properly, it is the practical mechanism that keeps a protected result on the correct side of the line when your notes reach an underwriter.

Because the family-history questions do the heavy lifting, and because a company-paid policy can quietly lift your total cover over the one threshold that matters, it is worth understanding how separate policies aggregate, which we set out in how a relevant life policy stacked on personal cover can cross £500,000. And if any of this leaves you wanting to see real figures for your own situation, you can get life insurance quotes and have an FCA-authorised broker match a policy to your circumstances.

Scotland and Northern Ireland

The predictive-versus-diagnostic distinction is set by the UK-wide Code, so the classification of any given test is the same across all four nations, and the £500,000 Huntington's threshold does not move at the border. What differs is the clinical apparatus around a test — how genetics services are organised, and, for the separate right to see a GP's report before it is sent to an insurer, the fact that Northern Ireland relies on its own 1991 Order rather than the Great Britain legislation. If you want the wider picture of how inherited risk and the odds fit together, our guide to family history, inherited risk and the odds is the place to start.

Frequently asked questions

What is the difference between a predictive and a diagnostic genetic test?

A predictive test looks for an inherited fault in someone who has no symptoms, in order to estimate future risk. A diagnostic test is done to explain symptoms or to confirm a condition you already have. Only predictive results are protected from insurers; a diagnostic result is ordinary medical evidence and is disclosable if you are asked about your health.

Which one is a cascade test after a relative's diagnosis?

Predictive, in almost every case, because you are being tested on the strength of the family finding rather than because you are unwell. That means the result cannot be asked for on a UK life application. The exception concerns research findings, which the Code handles separately, and any clinical follow-up the result triggers, which is disclosable.

Does it matter whether the test was NHS or private?

No. The Code treats a predictive result the same whether it came from an NHS clinical genetics service, a private clinic or a consumer kit bought online. What matters is why the test was done, and whether you had symptoms at the time, not who carried it out. The route the test took makes no difference to the classification.