What MOT mileage can—and cannot—tell you.

Start with the official record.

GOV.UK’s MOT history service includes mileage recorded at tests. Its historical coverage differs: car, motorcycle and van tests in Great Britain are available from 2005, while Northern Ireland results are available from 2017. The service describes further differences in certificate access. [1]

Choose two dated observations for the same vehicle. Confirm the unit, chronological order and whether either appears to be a duplicate test or retest. Do not quietly replace a suspicious value with one that produces a convenient result; flag it and investigate.

Use the actual number of days.

Subtract the earlier reading from the later one. Divide that distance by the elapsed days, then multiply by 365.25 for an annualised rate. Keep the measured distance and the annualised rate separate: the latter is a rescaling, not an additional observation.

An error is not proof of zero use or fraud.

DVSA provides a procedure for correcting mileage mistakes. Its current guidance distinguishes tests within the last 28 days from older records; older corrections require supporting evidence. Use the applicable official procedure rather than assuming a historical record is immutable. [2]

A decreasing reading, inconsistent units or an unusually large jump deserves investigation. Possible explanations are not conclusions. Keep evidence of the discrepancy and ask the provider what it accepts if the issue affects an insurance application. Do not upload private documents to this directory’s correction tool.

Then build a forward-looking estimate.

Suppose the observed period was before a change to home working, or included a long trip that will not recur. The arithmetic can be correct while the forecast is wrong. Rebuild the coming year from expected journeys and use the historical rate to explain the difference, not to override it.

For a newly acquired car, the previous keeper’s driving is not your own expected use. Similarly, the odometer covers the car’s journeys rather than one named driver’s miles. Record which period and whose circumstances each estimate reflects.