How many miles is too many on a used car?

Where the 7,100-mile figure comes from

It is the Department for Transport’s National Travel Survey, table NTS0901a, for 2024. It measures average miles per car per year, which is the number you want when you are looking at an odometer, rather than miles travelled per person, which is a different and smaller series that is easy to confuse with it.

Average annual mileage by fuel type

High mileage means something different for a diesel than for a petrol car, because the two are bought for different jobs and driven accordingly. The same reading on the same 10-year-old car is above average on one fuel and below it on another.

The gap between petrol and diesel is over two thousand miles a year, which compounds. Across 10 years it is roughly 21,000 miles of difference in what counts as ordinary. Applying a petrol expectation to a diesel is how a perfectly normal car ends up looking like a hard-worked one.

Working out the expected mileage for a specific car

Do not judge a number in the abstract. Turn it into a comparison in four steps.

1. Take the date of first registration from the V5C, not the model year in the advert.
2. Take the fuel type.
3. Multiply the age in years by the average for that fuel from the table above.
4. Compare that with the odometer, and treat a large gap in either direction as a question to ask rather than a verdict.

Read it as an average, not a ceiling. It is a mean, not a midpoint, so plenty of ordinary cars sit above their row and plenty sit below it, and a car well above the benchmark is not disqualified by that alone. What the comparison does is tell you which questions to ask next.

Why very low mileage is not automatically good

A car well under the benchmark for its age has spent that time doing something, and the usual answer is short journeys or long periods parked. Both cost something, and neither shows on the odometer.

• Short journeys. The engine and exhaust never reach full operating temperature. On a diesel that is the classic case in which the particulate filter does not regenerate properly, and on any engine it means more of the wear happens during cold running.
• Service items with a time limit as well as a mileage limit. Cambelts are the obvious one: manufacturers specify an interval in years as well as in miles, and a car that covers almost nothing can be years overdue while looking barely used. The same applies to brake fluid and coolant.
• Parts that age on the calendar. Tyres, rubber hoses and seals perish with time rather than with distance, and brake discs corrode while a car stands.
• A gap in the record. A car that has stood for a long time may have missed services and tests, and the reason it stood is a fair question to put to the seller.

None of this makes a low-mileage car a bad buy. It makes the service dates the thing to read, rather than the odometer. A low-mileage car with the belt done on schedule and a clean record is exactly what it appears to be.

Ten years old is the fleet average, not the old end

The average licensed car in the UK is 10 years old, on Department for Transport vehicle licensing statistics for the end of 2025. A car of that age is therefore ordinary rather than exceptional, and age on its own carries no more verdict than mileage does. What changes with age is that the evidence gets longer: more test results, more advisories, more invoices, more to read before you commit.

What matters more than the number

Mileage is the cheapest thing to look at and the least informative. Three other things tell you more, and all three are free to check.

• Service history. Stamps, and better still the invoices behind them. Look for the expensive scheduled work being done on time rather than for an unbroken run of oil changes.
• The pattern of MOT advisories. An advisory is not a failure; it is a note that something is wearing. The same advisory repeated across several years is the useful signal, because it shows something was flagged and left. See what an MOT advisory means, and the categories cars actually fail on in why cars fail the MOT.
• Evidence the mileage is genuine. Covered in the next section. A reading you cannot corroborate is not a low mileage; it is an unknown one.

DVSA recorded an initial failure rate of 28.08% for classes 3 and 4 in 2024/25. That is a national rate across every initial test in Great Britain, not the chance that the car in front of you will fail.

Checking the mileage is genuine

A mileage reading is taken at every MOT test and published free, which gives you a dated sequence to check the odometer against. This is the single most useful thing you can do with a registration number before you view a car.

• Readings should rise steadily. A reading that goes down needs an explanation.
• A long flat period followed by a jump is worth asking about, and so is an annual figure far out of line with the rest of the car’s life.
• Cross-check the mileages written on service invoices and stamps against the same sequence. Two independent records agreeing is much stronger evidence than either alone.
• The record starts at the first test, which is the third anniversary of registration in Great Britain and four years in Northern Ireland, so the earliest years of any car are not covered.

If the odometer and the published record cannot be reconciled, the car may not be as described. The Consumer Rights Act 2015 gives a short-term right to reject within 30 days when goods are not as described, and that applies to a car bought from a trader anywhere in the UK. Either way, a mileage you cannot corroborate is a reason to walk rather than a price to negotiate.

The mileage record itself is free from the GOV.UK MOT history service, which you search with the registration number; the MOT check tool here is a different thing, working out from the dates when a car’s next test falls due.