Nearside or Offside? What 91.6 Million MOT Defects Show

A UK car is not symmetrical in the way it meets the road. The left-hand side — the nearside — is the one that meets kerbs when parking, drain covers, the broken edge of the carriageway and the potholes that form there first. The right-hand side is the one the driver sits on and touches every day.

You might expect that to show up as a general excess of defects on the left. It does not. Across 91,592,223 defect lines recorded on 20,555,638 vehicles in 2025, the nearside carries 30,805,048 and the offside 30,661,257 — a difference of 0.47%. Split by corner it is flatter still: front nearside to front offside is 0.996, rear nearside to rear offside 1.015. On the average, the two sides of a British car are the same.

The interesting result is what happens when you stop averaging.

The wheel and the seat belt

Take the two mechanisms at the extremes, because between them they make the case that the side marker is measuring something real rather than a habit of how testers type.

A badly distorted road wheel is the closest thing the MOT vocabulary has to a record of a single hard impact — a kerb strike, or a pothole hit at speed. It is logged on the nearside 1.97 times as often as the offside: 322,209 against 163,859, from 486,068 sided lines.

Seat belt webbing damage runs the opposite way by an even larger factor — 88,618 offside against 38,006 nearside, 2.33×, from 126,624 sided lines. There is nothing mysterious about that: it is the driver’s belt, and it is pulled across a body on every journey while the front passenger belt often is not.

That mirror image is the whole point, and it is why the flat average matters rather than embarrasses. If the side marker were noise, both would sit near 1.00 like the fleet total does. If testers simply defaulted to one side, both would lean the same way. Instead the defect caused by the road skews hard to the road edge, the defect caused by the occupant skews hard to the seat that is always occupied, and the two effects cancel almost exactly when you add up every line in the country. The marker is tracking mechanism, and averaging destroys it.

Which mechanisms lean which way

Ranking every DVSA-coded defect item with at least 40,000 sided lines by its nearside-to-offside ratio produces a coherent ordering rather than a scatter. These are the items that lean to the kerb side.

Two of these are road exposure. A badly distorted wheel (1.97×, n = 486,068) and a tyre with a structural bulge (1.78×, n = 475,458) are the same event seen from the metal and the rubber: a sidewall and a rim pinched against the edge of something hard. Two independently coded items, the same story, the same side.

The item at the top of the table is not damage at all, and it is worth being clear about that. “Child Seat fitted not allowing full inspection of adult belt” (RfR 10003) is the tester noting that a fitted child seat stopped them inspecting an adult belt, and it runs 2.30× nearside on 725,040 sided lines. That is a fact about where families put child seats — behind the front passenger, on the kerb side, which is the side a child is lifted out on — rather than about wear. We include it because it is a second, completely separate confirmation that the coded side tracks how a car is actually used, and we label it for what it is rather than letting it pass as road damage.

And these are the items that lean to the driver’s side.

This list is use and corrosion rather than impact: both seat belt webbing items, brakes binding, a fractured coil spring, corroded rigid brake pipes. Nothing here is caused by the road edge, and nothing here leans to the kerb.

One item we excluded on purpose

The most one-sided item in the entire dataset is not in either table. “fog lamp not working” (RfR 31043) is recorded 83,919 times on the offside against 1,191 on the nearside — a ratio of 70.46 to one, far beyond anything wear produces.

It is excluded because it is not a finding about damage at all. UK lighting regulations require a single rear fog lamp on the offside or centre line, so a nearside rear fog lamp usually does not exist to be faulty. A defect can only be recorded on a component that exists. Leaving it in would have handed us a spectacular number describing nothing but the lighting regulations, and we mention it here rather than deleting it quietly because the same trap is waiting in any left-right analysis of vehicle data.

By component group

Rolling the same lines up to component groups shows the same split at a coarser grain — and shows how completely it disappears into the total.

Road wheels is the most nearside-skewed group at 1.743 — 405,068 nearside against 232,365 offside — and seat belts and restraints the most offside-skewed at 0.617, the same mirror image one level up. Tyres lean gently to the kerb (1.059) and brakes to the driver (0.888). Yet the two largest groups are near-symmetric, and they dominate the count, which is why the fleet total lands at 0.47% and tells you nothing.

The most-recorded MOT defects in Britain

Separately from the left-right question, the dataset answers a simpler one: what do MOT testers actually write down? These are the twenty most-recorded defect items, with how many of each are advisories rather than failures.

Tyre tread depth is far and away the most-recorded item in the country at 10,702,929 lines, of which 90.6% are advisories — a tread approaching the limit rather than already below it. Our tyre tread and pressure guide covers what the tester is measuring and how to check it yourself.

Most defects are advisories

This is the single most misread thing about MOT defect data. An advisory is a warning about something that will need attention before it becomes a failure — it is not itself a reason to fail, and a car can be issued a pass certificate carrying a long list of them. 66.7% of every defect line recorded in 2025 was one. At the other end, 3,543,478 lines were marked dangerous, across 2,531,499 tests — 6.09% of every class 4 and 7 test carried out in the year. Our explainer on advisories versus minor defects sets out what each category actually obliges you to do.

What we would do next, and cannot yet

The obvious follow-up is to test the road-surface explanation directly: take nearside wheel and tyre defect rates by area and compare them against local highway authority road-condition statistics. We cannot do it. The open dataset carries no postcode, so there is no geography to join on. Rather than gesture at a correlation we have not measured, we are stating the limit. If you hold or can point us to MOT records with an area marker, we would like to run it — hello@motcheckup.co.uk.