Data study · what drivers do next
MOT advisory statistics
The MOT advisory statistics nobody publishes are not about which faults appear, but about what happens afterwards. Across 138.5 million advisories followed forward, a third were never repaired at all — and the ones drivers skip are not the ones that matter least.
DVSA test recordOpen Government LicenceNo sign-up
33.4% of MOT advisories are never repaired, according to DVSA MOT data analysed by PriceMyRepair — not deferred, never, across every later test the vehicle takes. Fluid leaks are skipped most often at 48.6% on older cars, and they come back as a failure 1.3% of the time. Suspension advisories are skipped nearly as often and come back 18.9% of the time.
33.4%
of advisories are never repaired, across 138.5 million followed through the national test record
14.5×
how much the consequences vary between categories — against only 2.9× for how often they are skipped
26.6→38.0%
how the skip rate climbs from cars under seven years old to those over twelve
A third of every warning ever written
In PriceMyRepair’s analysis of 138.5 million DVSA MOT advisories, 33.4% were never repaired, and the share rises from 26.6% on cars under seven years old to 38.0% on cars over twelve. An advisory is a note a tester writes while a car is still passing. Nothing obliges anyone to act on it. This is what people actually do.
33.4% are never repaired. Not repaired late, not repaired at the next test — never recorded as resolved across the whole of that vehicle’s remaining history in the record. On 138.5 million advisories, that is roughly 46 million warnings written and permanently ignored.
The rate rises steadily with the age of the car. Under seven years old, 26.6% of advisories go unrepaired. Between seven and eleven, 32.0%. Past twelve, 38.0%. Nothing about the components changes across those bands to justify that — what changes is how much the car is worth relative to the work.
And the risk moves the same way. Conversion — the share of advisories that come back as an actual failure — runs 7.5%, 10.8% and 12.8% across the same three bands. So the older the car, the more likely both that a warning will be ignored and that ignoring it will cost something.
| Age of car | Never repaired | Came back as a failure | Advisories |
|---|---|---|---|
| Under 7 years | 26.6% | 7.5% | 30,930,934 |
| 7 to 11 years | 32.0% | 10.8% | 47,613,351 |
| 12 years and over | 38.0% | 12.8% | 59,935,031 |
Age bands are recorded in the source data rather than derived, so this is the one comparison on the site that needs no adjustment for fleet age. The three bands sum exactly to the total.
Drivers do not skip the cheap ones — they skip the invisible ones
Across twelve advisory categories, how often drivers skip a warning varies 2.9× while how often it comes back as a failure varies 14.5×, according to PriceMyRepair’s analysis — a correlation of only −0.351. The obvious theory is that people ignore what does not matter. Test it against the consequences and it only half holds.
Some of the skipping is entirely rational. Fluid leaks are the most-skipped category of all, and they are also the least consequential: 1.3% come back as a failure. A driver who reads “slight oil leak, not excessive” and does nothing has made a decision the data supports.
But the second most-skipped category is the most consequential one on the list. Suspension advisories are ignored at almost the same rate as fluid leaks and come back as failures 18.9% of the time — fourteen times more often. Nothing in the wording of a certificate distinguishes the two, and the skip rates say drivers cannot tell them apart.
Measured across all twelve categories, the relationship is weak. The correlation between how often a category is skipped and how often it returns as a failure is −0.351: real, but far from the −1.0 that perfect judgement would produce. What people are sorting on is how alarming the phrase sounds, not what happens next.
The scale of the mismatch is the point. Skip rates vary 2.9× across categories, from 15.3% to 43.8%. Consequences vary 14.5×, from 1.3% to 18.9%. Drivers are making a narrow range of decisions about a very wide range of risks.
Every category, and what skipping it is worth
Share never repaired, broken out by the age of the car when the advisory was written, with the failure rate for that category alongside.
| Category | Under 7 | 7–11 | 12+ | Becomes a failure | Advisories |
|---|---|---|---|---|---|
| Fluid leaks | 25.3% | 37.7% | 48.6% | 1.3% | 8,483,660 |
| Vehicle identification | 30.2% | 38.0% | 44.8% | 3.8% | 3,962,778 |
| Suspension | 28.7% | 35.0% | 40.9% | 18.9% | 26,395,027 |
| Brakes | 26.7% | 35.1% | 44.1% | 9.5% | 35,121,620 |
| Other equipment | 24.7% | 31.2% | 33.1% | 6.7% | 1,053,047 |
| Tyres | 27.1% | 30.8% | 34.4% | 10.0% | 45,227,178 |
| Body and exhaust | 20.6% | 24.8% | 30.7% | 12.1% | 5,970,177 |
| Headlamps | — | 19.3% | 26.2% | 15.0% | 1,042,037 |
| Axles and bearings | 19.7% | 20.5% | 26.5% | 6.8% | 1,948,692 |
| Signals | 15.9% | 20.3% | 23.5% | 8.2% | 1,685,625 |
| Steering | 16.8% | 17.6% | 23.6% | 15.8% | 2,420,117 |
| Brake performance | 9.9% | 14.1% | 17.7% | 9.6% | 3,731,937 |
Headlamps carry no published figure for cars under seven years old, because the sample at that age is too small to report; the cell is left empty rather than filled with a zero. Categories below one million advisories are excluded entirely. Shaded rows mark the three categories where skipping and consequence are furthest apart in either direction.
Suspension advisories turn into MOT failures 18.9% of the time and steering advisories 15.8%, according to PriceMyRepair’s analysis of DVSA data — the highest and third-highest on the table — and both sit in the middle or upper half of the skip rankings. Between them they account for nearly 29 million advisories.
The reason is probably that neither announces itself. A worn bush or a joint with slight play changes nothing a driver would notice from the seat, so the certificate is the only evidence there is, and a phrase like “slight play” reads as reassurance rather than warning. Brake performance, by contrast, is skipped least of anything here — the word “brake” paired with the word “efficiency” does the work that “slight play” does not.
Our companion study ranks the same categories by what each one costs when it does come back, which is the other half of this picture.
As cars age, everyone stops fixing things
Every one of the 22 manufacturers in PriceMyRepair’s data shows a rising skip rate with vehicle age, and the gap between them narrows from 1.75× on cars under seven to 1.36× on cars over twelve. Split the same figures by manufacturer and one pattern survives scrutiny while another does not.
Every manufacturer’s skip rate rises with age. All twenty-two of them, without a single exception. That consistency is unusual in this dataset and it says the effect is about ownership economics rather than any property of the vehicles.
The differences between manufacturers shrink as cars get older. Among cars under seven, the gap between the highest and lowest skip rate is 1.75×. Between seven and eleven it narrows to 1.48×. Past twelve years it is 1.36×. Whatever separates one badge’s owners from another’s early on, it fades.
What we will not publish is a ranking of manufacturers. We built one, then tested it, and it failed: the order at twelve-plus years does not reproduce among younger cars. A badge that appears near the top of the older group mostly reflects that cars wearing it survive long enough to be in that group at all, which measures longevity rather than what owners choose to do. The shape of the change is robust; the league table is not, so it stays out.
What this means for the certificate in your glovebox
A third of advisories going unrepaired is a fact about the country. What it should change for one driver is narrower and more useful.
Sort by category, not by wording
The certificate’s language does not signal risk. A suspension note reading “slight play” is fourteen times likelier to end in a failure than an oil leak “not excessive”, and neither sounds urgent.
The category is the signal, not the adjectiveRepeats matter more than firsts
An advisory that appears on consecutive tests is one that was skipped and is still there. That is the pattern this whole study is made of, seen on a single vehicle.
Two years running is a decision, not a noteSkipping is sometimes right
A fluid leak on an older car converts at 1.3%. Choosing not to act is a defensible call, and this page is not an argument for fixing everything.
Some advisories genuinely are just notesThe single most useful thing on any used car listing is its test history, which names every advisory a tester has ever recorded and shows which ones came back. Our guide to what the MOT actually checks sets out what the test can and cannot see.
Enter your reg — we’ll read the full test history, translate every advisory a tester has written and price the repair for your model rather than the national average.
How these figures were produced
Method
Source. The DVSA MOT testing bulk dataset, published under the Open Government Licence v3.0 and analysed by PriceMyRepair, covering Great Britain. Northern Ireland runs a separate system under the DVA and is not in this data.
What “never repaired” means. Each advisory was followed forward through every subsequent test on the same vehicle. An advisory counts as never repaired when no later test records it as resolved — the item either persists or the vehicle stops appearing in the record. It is the complement of the repaired and the converted-to-failure groups, not a separate judgement.
Age bands are in the source, not derived. The three bands — under seven, seven to eleven, twelve and over — are recorded against each advisory as written. That removes the fleet-age problem that distorts most model-level comparisons, and it is why this page leads with bands rather than with a league table.
Why the total is 138.5 million and not 145.1 million. Our other studies cite 145,100,479 advisories. This one covers 138,479,316, and the gap is fully accounted for. Four defect categories present in the national totals are absent from the model-level table and contribute 51,234 — under one per cent of the difference. The remaining 6,569,929, or 4.53% of the whole, are advisories on vehicles whose make and model could not be resolved to a single canonical name, and which therefore cannot appear in a table built per model. Nothing has been excluded on judgement.
Correlation. Pearson’s r between the skip rate and the failure rate across the twelve categories shown: -0.351. Skip rates range 15.3% to 43.8%; failure rates 1.3% to 18.9%.
Known limitations. A vehicle leaving the record — scrapped, exported, or off the road — is indistinguishable from one that kept an advisory unrepaired, so the skip rate at twelve-plus years is an upper bound. An advisory and a later defect in the same category are treated as the same fault, which slightly overstates conversion where a different component in that group failed instead. Manufacturer figures are shown as a shape rather than a ranking for the reason set out above. Headlamps carry no figure below seven years because the sample is too small to publish, and categories below one million advisories are excluded.
Contains public sector information licensed under the Open Government Licence v3.0.
MOT advisory statistics — FAQ
How many MOT advisories are never repaired?
33.4%, across 138.5 million advisories followed through the national test record. That is roughly 46 million warnings written by testers and never acted on. The share rises with the age of the car, from 26.6% on cars under seven to 38.0% on those over twelve.
Which MOT advisories do drivers ignore most?
Fluid leaks, at 48.6% on cars over twelve, followed by vehicle identification items and brakes. Suspension is close behind, which is the finding that matters: it is skipped almost as often as fluid leaks while being fourteen times likelier to come back as a failure.
Is it a mistake to ignore an MOT advisory?
Not always, and the data is clear about it. A fluid leak recorded as not excessive converts into a failure 1.3% of the time, so leaving it is a defensible decision. The mistake is applying the same reasoning to suspension or steering, which convert at 18.9% and 15.8% and read no more alarmingly on the certificate.
Do older cars get more advisories ignored?
Considerably. The skip rate runs 26.6% under seven years, 32.0% between seven and eleven, and 38.0% past twelve, and every manufacturer in the data follows that curve without exception. Nothing about the components changes; what changes is the repair bill relative to what the car is worth.
Do drivers ignore the advisories that matter least?
Only loosely. The correlation between how often a category is skipped and how often it becomes a failure is −0.351 — real judgement, but weak. Skip rates vary 2.9× across categories while the consequences vary 14.5×, so a narrow range of decisions is being applied to a very wide range of risk.
Which manufacturer’s owners ignore advisories most?
We tested that ranking and will not publish it. The order among cars over twelve does not reproduce among younger ones, which means it largely reflects which badges survive long enough to appear in the older group. What is robust is the shape: every manufacturer’s skip rate rises with age, and the differences between them narrow from 1.75× to 1.36×.
What happens if an advisory is never fixed?
Usually nothing, for a while. Averaged across categories, 10.9% of advisories return as an actual failure at a later test, and the median gap between the note and the failure is about one MOT cycle. The remainder either persist harmlessly or leave the record with the vehicle.
Why does this page say 138.5 million when your other pages say 145.1 million?
Because this table is built per make and model, and advisories on vehicles whose make could not be resolved to a single canonical name cannot appear in it. Those account for 4.53% of the total. A further 51,234 sit in four defect categories absent from the model-level table. The two figures describe the same dataset cut differently.
Does a repeated advisory mean anything?
It means the first one was skipped and the fault is still there — this entire study, observed on a single vehicle. A note appearing on consecutive certificates has moved from being a warning to being a decision, and it is the clearest signal a test history offers about how a car has been kept.
Can I see which advisories my own car has?
Yes, and it costs nothing. Every advisory a tester has ever written is public and tied to the registration, with dates. Our advisory decoder reads that history, translates the wording into plain English and prices each repair for your model.
Related studies and guides
The same dataset cut along other axes.