How long does a check take?

The free check normally returns in a couple of seconds. Results for a registration checked recently are served from a short-lived cache, which makes them close to instant without ever showing you data old enough to have changed.

Why do you show your reasoning instead of a score?

Because the reasoning is the useful part. "Seven out of ten" tells you nothing you can act on. "The same corroded subframe advisory at three consecutive tests, getting worse each time" tells you what to ask about and what to budget for.

Report addresses are random references rather than registrations, so they cannot be guessed, and they carry a directive telling search engines not to index them. We do not publish who checked what.

How the check works

You enter a registration. We ask the DVLA vehicle register and the DVSA MOT record what they hold, read the result, and show you both what was found and how we reached each conclusion.

There is no score out of ten anywhere in this. A single number is easy to put on a page and impossible to argue with, which is precisely the problem: it hides the reasoning, and the reasoning is the part that helps you decide whether to walk away or negotiate.

What happens when you press the button

The free check makes two lookups, and both are against official government sources.

The DVLA vehicle enquiry service returns make, colour, fuel, engine size, year, CO₂, Euro status, tax status and MOT status.
The DVSA MOT history service returns every recorded test, its result, the mileage on the day, and the exact wording of every defect and advisory.
Everything else on the free report is worked out from those two results. The mileage analysis, the recurring-fault grouping, the corrosion tracking, the emissions-zone assessment and the road tax explanation all cost nothing extra, because they are analysis rather than another lookup.
The paid sections make one further lookup, against the finance, police and insurance registers, which charge per enquiry.

How we decide what a finding means

Every rule that reads the data is written down, versioned, and stored against the report that used it. If the analysis changes next year, an old report still records which version of the reasoning produced it.

The rules are also deliberately conservative in one direction. Where something could be either an ordinary recording error or a real problem, the ordinary explanation is tested first, because a check that reports every anomaly as fraud is one you learn to ignore, and then you ignore the real one too.

What happens when a source cannot be reached

Registers go down. When one does, the report says that section could not be checked, and it says it before anything reassuring.

It never shows a green result for something it did not manage to look at. That sounds obvious and it is not how this industry generally works: the ordinary pattern is a page of ticks where a tick means "no problem was returned", which is not the same as "no problem exists". If you paid for a check that did not run, you are not charged for it.

What we do not do

We do not show the names or addresses of keepers. Keeper history is dates only, which is all you need to judge whether a vehicle has changed hands unusually often, and none of the rest is ours to publish.

We do not put registrations in report URLs. A report address is a random reference, so it cannot be guessed and nobody can enumerate other people's checks by counting upwards.

We do not claim to assess mechanical condition. This reports records. It is not an inspection, and it does not replace one.

Common questions

How long does a check take?: The free check normally returns in a couple of seconds. Results for a registration checked recently are served from a short-lived cache, which makes them close to instant without ever showing you data old enough to have changed.
Why do you show your reasoning instead of a score?: Because the reasoning is the useful part. "Seven out of ten" tells you nothing you can act on. "The same corroded subframe advisory at three consecutive tests, getting worse each time" tells you what to ask about and what to budget for.
Is my check private?: Report addresses are random references rather than registrations, so they cannot be guessed, and they carry a directive telling search engines not to index them. We do not publish who checked what.