HSE tested asbestos laboratories blind. Here is what it found.
HSE sent previously validated proficiency samples to 59 laboratories, disguised as routine work from a doctor’s surgery. The headline error rates alarmed people. The detail tells a more balanced story, but it also shows exactly where laboratories need to improve, and what their clients will now ask them.
- Between October 2024 and February 2026, HSE sent four validated AIMS samples to 59 laboratories as if they were routine commercial work.
- Across 236 analyses there was one true false negative and no false positives: asbestos was detected in 99.6% of cases where it was present.
- Errors identifying every asbestos type in multi-type samples were far higher than in proficiency testing, reaching 76.3% on one cement sample.
- Only 7% of laboratories flagged the deliberately deviating sample, despite this being a requirement of HSG248 and ISO 17025.
- Clients will ask harder questions about quality control. Laboratories that explain theirs openly will win their trust.
What HSE did
The Asbestos in Materials Scheme (AIMS) has been the UK benchmark for laboratory proficiency since the mid-1990s. Participation is a precondition for UKAS accreditation of asbestos identification in the UK. Laboratories generally perform well in it. The question HSE wanted to answer was whether routine commercial work matches that performance.
Between October 2024 and February 2026, working through a third party, HSE sent four samples that had already been used in AIMS rounds to 59 laboratories. They were presented as materials from a doctor’s surgery ahead of refurbishment, with nothing to suggest HSE was involved. Because the samples had known AIMS results, HSE could compare the two directly. HSE’s Fibres and Minerals team has since presented the findings to the industry, including at a NORAC webinar promoted by UKATA.
The results
| Sample | Contained | AIMS error rate | Mystery shopper error rate |
|---|---|---|---|
| 1. Asbestos insulating board | Chrysotile and amosite | 0.6% | 25.4% |
| 2. Cement | Chrysotile, amosite and crocidolite | 35.1% | 76.3% |
| 3. Textured coating | No asbestos (deviating sample) | n/a | No false positives |
| 4. Vinyl floor tile | Chrysotile in tile and adhesive | 1.6% | 2% |
Figures as published by ATaC in the Autumn 2026 issue of ARCA and ATaC News.
Extrapolated across three AIMS rounds, 78% of laboratories would have fallen into the unsatisfactory category. ATaC points out that this is a stress test rather than a prediction, because it assumes the same samples and the same errors repeated with no quality control stepping in.
The reassuring part
Across 236 analyses there was one true false negative and no false positives. Asbestos was detected in 99.6% of the cases where it was present, which is arguably the measure that matters most to a duty holder. Every laboratory found the amosite in sample 1 and the chrysotile in sample 2. In most cases, missing the additional type would not have changed the material risk assessment or the management decision.
The part that needs fixing
Laboratories are meant to identify every asbestos type present, and too many did not. The pattern points to behaviour under commercial pressure rather than a lack of skill: analysts stopping once the first type is found, and amphiboles being overlooked where chrysotile is abundant. The other clear weakness was the deviating sample. Only 7% of laboratories recognised it, even though recognising and reporting deviations is required by HSG248 and ISO 17025.
Part of the gap is structural. Under UKAS requirements every analyst examines AIMS samples, so proficiency results reflect a laboratory’s best collective effort. In routine work, only around 14 to 16% of samples are typically rechecked. The study measured a single analyst’s everyday work, which is precisely what clients receive.
The UK is not alone. A 2026 study in the Journal of Occupational and Environmental Hygiene that sent blind bulk samples to laboratories across several countries also found false negative rates well above proficiency results in early rounds, falling as the programme continued.
Can a surveyor see your quality control before they send samples?
We build laboratory websites that explain accreditation, AIMS participation, recheck rates and how unusual samples are handled, in plain English, alongside turnaround and online sample submission.
What laboratories should do next
- Change the habit, not just the procedure. Train and audit analysts to keep searching after the first asbestos type is identified.
- Make deviating samples a standing item in training and internal audit, with clear reporting routes.
- Review recheck rates against risk. Multi-type materials such as cement and insulating board deserve more scrutiny than single-type materials.
- Treat every missed type as a learning event, whatever its effect on the risk assessment.
- Explain your quality control to clients in plain English, because they are now asking.
What clients will now ask
Surveyors, consultants and contractors read the coverage too. Expect questions about recheck rates, how analysts are trained on multi-type materials and how unusual samples are handled. Duty holders and their advisers also have a sharper reason to care: HSE has said it will hold third-party advisers to account where poor advice leads to a breach, which we explain in our guide to sections 36 and 37.
Most laboratory websites cannot answer those questions. A UKAS logo tells a client you are accredited, not how you work. Laboratories that publish a clear quality control page, state their turnaround honestly and make samples easy to send will win the cautious clients. We explain the wider principles in how to show asbestos credentials online and competence, accreditation and certification explained, and we track every development on our regulation and policy tracker.
We have over 20 years of experience in web design and digital marketing for the trades, and we build websites for asbestos laboratories using data, not guesswork. Start with a free Four Stage Clearance.
Sources: ATaC analysis of the HSE AIMS mystery shopper study, ARCA and ATaC News Issue 133, Autumn 2026; HSE AIMS information book; UKATA; Journal of Occupational and Environmental Hygiene, 2026.
Questions about the mystery shopper study
What was the HSE asbestos mystery shopper study?
Between October 2024 and February 2026, HSE worked through a third party to send four previously validated AIMS proficiency samples to 59 laboratories across Great Britain. The samples were presented as routine materials from a doctor’s surgery before refurbishment, with no reference to HSE.
Did laboratories miss asbestos?
Rarely. Across 236 analyses there was one true false negative and no false positives, a 99.6% success rate in detecting asbestos where it was present. Most errors involved missing a second or third asbestos type in samples containing more than one.
Why did laboratories perform worse than in AIMS?
AIMS samples are known to be proficiency samples and, under UKAS requirements, are examined by every analyst in a laboratory. Routine samples are usually examined by one analyst, with only around 14 to 16% rechecked. The study measured routine practice rather than best-case capability.
What is a deviating sample?
A sample that does not match what it claims to be or that raises a question about its suitability for analysis. Laboratories are expected to recognise and report these under HSG248 and ISO 17025, but only 7% did so in the study.
What should asbestos laboratories do now?
Strengthen procedures and training so analysts keep searching after finding the first asbestos type, improve recognition of deviating samples, review recheck rates against risk and explain quality control clearly to clients.
Make your quality visible.
Run the free Four Stage Clearance and see whether your laboratory website shows surveyors and contractors the accreditation, quality control and turnaround they now look for. Our team will then personally review it within one working day.
