Calculated from official DVSA data

MOT pass rates by make and model

Every MOT in Britain goes on record. We've crunched the DVSA's 2025 results for the 36 makes we cover, 32,727,686 car and light van tests or 98.5% of the year's total, into honest pass rates for the makes and models people actually buy. No survey, no opinions, just what happened in the test lane. For the models that fail least for their age, see the most reliable used cars.

Tests on the 36 makes covered: 32,727,686
Failed first time: 27.7%
Data year: 2025

Pick a make

Initial fail rate = tests that failed, plus passes that needed a defect fixed on the spot. Lower is better.

Make	Tests in 2025	Initial fail rate
Abarth	30,296	22.0%
Alfa Romeo	67,745	32.1%
Audi	1,762,872	21.5%
BMW	1,887,439	20.7%
Citroen	985,721	36.5%
Dacia	218,232	29.9%
DS Automobiles	47,223	31.2%
Fiat	780,698	32.3%
Ford	4,366,490	31.2%
Honda	899,468	27.5%
Hyundai	956,409	26.1%
Isuzu	45,749	31.6%
Jaguar	348,228	20.6%
Jeep	71,149	27.4%
Kia	1,030,526	24.8%
Land Rover	943,188	22.7%
Lexus	170,430	16.0%
Mazda	534,114	27.3%
Mercedes-Benz	1,743,653	22.9%
MG	171,040	19.1%
MINI	821,305	22.6%
Mitsubishi	273,514	28.7%
Nissan	1,666,188	32.2%
Peugeot	1,395,100	33.2%
Porsche	191,793	12.5%
Renault	954,670	37.5%
SEAT	561,947	26.0%
Skoda	813,597	23.6%
Smart	85,097	28.7%
Subaru	57,336	25.5%
Suzuki	436,545	26.4%
Tesla	139,546	16.2%
Toyota	1,749,337	24.9%
Vauxhall	2,816,986	34.2%
Volkswagen	3,098,837	27.0%
Volvo	605,218	24.7%

How we work these figures out

The DVSA publishes an anonymised record of every MOT: the make and model, the car's age, the result and what was wrong. We take the complete 2025 release for cars and light vans (class 4 tests), keep only normal tests, and count a test as an initial fail if it failed outright or passed only after a defect found in the lane was put right.

Model names in the raw data are messy. The same car appears as "COOPER", "COOPER S" and "MINI", and a BMW 3 Series hides behind badges like 320D. We fold those variants into the model family a buyer would search for, and we only publish a figure when a model has at least 12,000 tests in the year. Small samples make impressive-looking nonsense, and we'd rather show fewer numbers you can trust.

One honest caveat. These are averages across every example of a model, kind owners and careless ones alike. The car you're actually looking at has its own record, which is exactly what our free MOT check and the £5.00 report show you. Statistics say what usually happens. The history says what happened to this one. Our MOT history guide covers how to read it.

Contains DVSA anonymised MOT testing data © Crown copyright, licensed under the Open Government Licence v3.0.

Figures last checked against the official source on 18 August 2026. Check the source yourself.

MOT statistics: common questions

Where do these MOT statistics come from?

From the DVSA's anonymised MOT test results, the official record of every test carried out in Great Britain. We process the full 2025 data, over thirty million car and light van tests, and publish rates only where a model has enough tests for the numbers to mean something.

What counts as a fail in these figures?

We quote the initial fail rate: tests that ended in a fail plus those that passed only after a defect was fixed during the visit (PRS). It measures the state cars arrive in, which is what matters to a buyer. A rate counting only outright fails would look flatteringly low.

Does a low fail rate mean a car is reliable?

It means the model tends to arrive at its MOT in good order, which reflects build quality as well as how its owners treat it. The MOT doesn't test the gearbox, clutch or electronics, so read these rates alongside a specific car's own MOT history rather than instead of it.