Diesel vs petrol reliability

The same number, for a quarter more miles

Set petrol and diesel next to each other at matched ages and the headline is that nothing separates them. That headline is true and almost useless.

The failure rates converge and then sit on top of each other. At three years old diesel is worse — 17.6% against 14.3%. By six years the gap has closed to 1.5 points. At ten years diesel is fractionally better, 30.2% against 30.5%. At fourteen the two are identical to one decimal place.

What differs is the odometer behind each figure. A three-year-old diesel has covered 41,272 miles against 27,639 for petrol — 49% further. The gap narrows but never closes: 36% at six years, 26% at ten, 28% at fourteen. Diesels are bought to be driven, and they are.

The components an MOT inspects wear on distance, not on time. Tyres, brakes, suspension bushes, drop links, discs — none of them care how many birthdays a car has had. So comparing two fuels at matched ages while ignoring the mileage compares two things that were never asked to do the same job.

Per mile, diesel wins at every age

Divide each failure rate by the miles behind it and the picture inverts. Compare only down a column, never across — the note below explains why.

Diesel’s per-mile advantage over petrol runs 18%, 21%, 22% and 22% across four, six, ten and fourteen years — remarkably stable across 60 million tests. Read this column down, not across. The mileage figure is lifetime, not annual, so the per-mile rate falls with age as an arithmetic certainty rather than a finding: between three and fourteen years, petrol failure rises 2.3× while mileage rises 3.3×. Comparing fuels within one age is sound; comparing one fuel’s figure across ages is not, and we do not.

Hybrids need no adjustment at all

Everything above depends on a derived metric, and derived metrics invite argument. This next finding does not need one.

At every age, hybrids cover more miles than petrol cars and fail less often. Both things at once, in the same direction, with no division and no normalisation. At three years: 20% further, 4.1 points lower. At six: 28% further, 8.5 points lower. At ten: 24% further, 11.6 points lower. At fourteen: 35% further, 9.2 points lower.

That is dominance rather than a ranking. No adjustment for use can reverse it, because the harder-used group is also the better-performing one. It is the only comparison on this site where the answer survives every objection we could think to raise against it.

Electric sits between hybrid and diesel and the sample is thinner. 23.8% at ten years on 93,957 miles, from 61,240 tests — enough to report, not enough to lean on. At fourteen years the electric fleet drops to 26,317 tests and we would not publish a finer breakdown than this from it.

Engine size runs backwards to the raw numbers

The same correction applied to engine capacity produces an ordering the raw failure rates hide completely.

All figures at ten years old. Raw failure rates sit in a narrow band between 25.3% and 31.3% and say almost nothing; per mile the ordering is clean and monotonic from 1.0 upwards. Vehicles with no recorded engine capacity are excluded.

The smallest engines break the pattern, and the reason is worth naming. Cars under 1.0 litres have the worst per-mile figure of all at 4.80, despite the lowest raw failure rate. They also cover by far the fewest miles — 54,559 against 97,076 for a 1.6. These are city cars doing short, cold, stop-start journeys, and that is the hardest kind of mile there is for brakes and suspension. Miles are not interchangeable, which is the limit of this whole method.

What an MOT can and cannot say about a diesel

This is a study of one test, and the test has a shape.

This distinction matters most for diesels, where the components outside the test are both expensive and characteristic. See what the MOT actually checks for the full boundary, and DPF costs for the largest of the items it cannot see.

How these figures were produced

Diesel vs petrol reliability — FAQ

Are diesel cars more reliable than petrol?

On MOT evidence, per mile driven, yes — by 18% to 22% at every age measured. At matched ages the raw failure rates are almost identical, but diesels have covered 26% to 49% more miles to reach them. The test does not inspect turbochargers, injectors or particulate filters, so this is a statement about wear items rather than about ownership costs.

Do diesels fail their MOT more often than petrol cars?

Only when young. At three years diesel fails 17.6% against petrol’s 14.3%, a gap that closes by six years and disappears entirely by ten. At fourteen years both sit at 33.2%. The early gap tracks the mileage gap, which is widest in the first years of a diesel’s life.

How many more miles does a diesel do than a petrol car?

At three years old, 41,272 against 27,639 — 49% further. The gap narrows with age but never closes: 36% at six years, 26% at ten and 28% at fourteen. A fourteen-year-old diesel has typically covered 117,895 miles.

Are hybrids more reliable than petrol cars?

By the clearest margin in this data. At every age hybrids cover more miles than petrol cars and fail less often — both at once, needing no adjustment for use. At ten years a hybrid has done 24% more miles and fails 11.6 percentage points less. Regenerative braking is the likeliest mechanism, since it removes much of the work from friction brakes.

How reliable are electric cars at MOT?

Better than petrol and diesel, behind hybrids. At ten years old an electric car fails 23.8% of tests on 93,957 median miles. The samples are thin — 61,240 tests at ten years and 26,317 at fourteen — so we report the headline and publish no finer breakdown from them.

Do bigger engines fail their MOT less often?

Per mile, consistently. From 1.0 litres upwards the rate falls monotonically: 4.28, 3.68, 3.20, 2.78 and 2.65 failures per ten thousand miles. Raw failure rates hide this entirely, sitting in a narrow band between 25% and 31% because larger-engined cars cover far more miles.

Why do sub-1.0-litre cars do worst per mile?

Because their miles are the hardest kind. They cover 54,559 miles by ten years against 97,076 for a 1.6, and those miles are short, cold and stop-start — the pattern that wears brakes and suspension fastest. It is the clearest reminder that a mile is not a fixed unit of wear.

Does this mean a diesel is cheaper to run?

No, and the data cannot answer that. The MOT does not inspect particulate filters short of visible smoke, turbochargers, injectors, EGR valves or dual-mass flywheels — the components that produce the expensive, characteristically diesel bills. A car can pass every test it takes and still cost four figures.

Why not compare the per-mile figures between different ages?

Because the mileage in the denominator is lifetime rather than annual, so the figure falls with age whatever the cars are doing. Between three and fourteen years petrol failure rises 2.3× while mileage rises 3.3×, which makes the decline arithmetic rather than evidence. Within a single age the denominator is shared and the comparison holds.

Should I buy a diesel or a petrol used car?

This data says a used diesel has usually done more work for the same MOT record, which is a point in its favour if the mileage does not put you off. It says nothing about the turbo, the filter or the injectors. Read the individual car’s test history rather than the fuel type — our used car check pulls it from the registration.

Related studies and guides

This page supplies the mileage evidence the other studies could only assume.