MOT Tests & Results (Anonymised)
Every MOT test in Great Britain since 2005 — make, model, age, mileage, pass or fail, and exactly what it failed on. The best free evidence of which vehicles hold up and which do not.
Motor trade & automotive Logistics Insurance Transport & travel
| Publisher | Driver and Vehicle Standards Agency |
|---|---|
| Licence | Open Government Licence v3 what this means |
| Updates | Updated annually |
| Cost | Free — open data |
| Last verified | 2026-07-16 |
What your business can use it for
Risk assessment
Before you buy a van or a fleet, check how that make, model and age actually performs at MOT — real failure rates and real failure reasons, across millions of tests.
Market analysis
Understand the vehicle parc and what goes wrong with it: which components fail, at what age and mileage — the demand signal behind any parts, servicing or warranty proposition.
Works with data you already have
-
Roughly where you operate — at region level, not street level
Postcodes
The only geography is a coarse postcode-area code for the test location; areas with fewer than five test stations are merged into a single suppressed code
How to get it
-
Bulk download
Anonymised MOT tests and results — 2023 onwards (DVSA)
No account needed. The actively maintained listing. One row per test, plus a companion file with one row per failure reason, linked by test id. -
Bulk download
Anonymised MOT tests and results — 2005 to 2023 (DfT archive)
The historic half of the same series. DVSA took over publishing from 2023, so you need both listings for the full run.
What's involved in using it
- For each year there are two big files — the tests and the reasons they failed — that you have to join together, plus several small code-lookup files to translate the numbers into readable labels, so it's an assembly job before any analysis.
- The files are large (several gigabytes per year), so they need database tooling rather than a spreadsheet, and each year is a separate download to stack.
- The only location is a coarse regional code for where the test happened — with sparse areas merged into a catch-all — so you can't tie a test to a specific place, station or customer.
Worth knowing before you start
- The geography is deliberately coarse — a postcode-area-level code for the test location, with thin areas merged into a catch-all. You cannot locate a test, a station or a customer from this.
- It is anonymised: the vehicle identifier is a synthetic id, not a registration, so you cannot look up a specific vehicle. Use the separate MOT History API for that — it needs registration.
- It runs well behind: the newest confirmed release is 2024 data, published in May 2025. Treat it as a historical series, not a current one.
- The 2018 MOT rule change altered how defects are recorded, so failure categories are not comparable across that boundary.
- Some years have been revised after publication (2017 reissued in 2022, 2022 corrected in 2023) — pin the version you analysed.
- MOTs skew to older vehicles: cars are not tested until three years old, so new-car reliability is invisible here.
Ask us about MOT Tests & Results (Anonymised)
Ask a factual question about MOT Tests & Results (Anonymised) — what it covers, how often it changes, whether it matches the records you keep — and we will answer, free. Or tell us what you are trying to work out, and we will tell you what it would take to get this data working next to the numbers you already have.