How we work out our MOT statistics

The data holds one record for every MOT test, with the vehicle's make, model, fuel, first-use date and recorded mileage, the test result and the postcode area of the test station, plus every failure, repair and advisory item. Vehicles and testers are anonymised. It contains public sector information licensed under the Open Government Licence v3.0. The analysis is Motoring Insider's; DVSA does not endorse it.

• Class 4 vehicles: cars, taxis, motor caravans, minibuses with up to 12 passenger seats and goods vehicles up to 3,000 kg. Motorbikes, buses and goods vehicles from 3,000 to 3,500 kg (class 7) are left out.
• Normal tests only. Retests after a failure are left out, as are aborted, abandoned and refused tests. That leaves 33,210,951 tests.
• Passed first time means the car passed outright. A car repaired during the test visit and then passed counts as not passing first time, because it had a fault on arrival.
• Age is 2025 minus the year the car was first used. Age comparisons use cars 3 years and over, the age of a car's first MOT.
• Failure items include faults fixed at the station. A car with two faults in one area counts once for that area.
• Models: DVSA records the model as written on the test, so "FIESTA ZETEC" and "FIESTA" are both a Fiesta. We group each make's model names into the models people know, and leave out names we cannot place.
• Mileage: recorded mileages that are missing or over 400,000 are left out. Typical means the median.
• Postcode area is where the car was tested, which is usually but not always where it is kept.

Older cars fail more often, so a make or model with mostly new cars will always look good on its raw pass rate. To compare fairly we work out what its pass rate would have been if each of its cars had passed at the national rate for its age.

"Vs same-age cars" is the actual pass rate minus that expected rate, in percentage points. The age-adjusted pass rate adds the same difference to the national pass rate for cars 3 years and over (72.2%), so every make, model and area is on one scale.

Make X's raw 83.0% looks well ahead of the 72.3% national figure, but most of that gap comes from having so many young cars. Allowing for age, it is 1.1 points better than cars of the same ages, and the age-adjusted rate puts that on the same scale as every other make.

A model's reliability score is its age-adjusted first-time pass rate. We only score models with at least 20,000 tests on cars 3 years and over, so a handful of unlucky or lucky cars cannot move the result.

An MOT checks safety and emissions on the day. It does not record breakdowns, repairs between tests or running costs, and it reflects how well cars are looked after as well as how they are built. A high score means a model passes its MOT more often than others of the same age; it cannot tell you about one car.

For the advisory figures we matched each car's last normal test in 2024 to its first normal test in 2025 (28,838,361 cars). For the 40 most common advisories, we give the share that did not pass the second test first time, against all matched cars of the same ages.