Same money buys a four year old car with 90k miles or a nine year old with 35k - which is actually the better bet?

The thing that decides more of these than any of the above, and which people inspect last: get underneath both cars.

Corrosion is what actually ends most cars in a wet climate, and it is a function of age, salt exposure and where it lived - not of mileage at all. A nine year old car has had nine winters. Sills, subframes, brake lines, suspension mounts, and the area around the rear arches. Surface rust is normal; flaking, bubbling under paint, or anything soft when prodded is a different matter.

Brake lines in particular are worth a look on any older car, they corrode from the outside and are an expensive and non-negotiable repair.

Things I would do before choosing:

Pay for an independent inspection on the one you prefer. It costs a fraction of one repair and it is the highest-return money in the whole process.

Check both against a history service for write-offs, outstanding finance and mileage discrepancies.

Look at what each will cost to insure and tax before you decide, since these can differ enough to change the maths.

And the honest summary: given the two you describe, I would look hard at the high-mileage four year old with good history first, and treat the nine year old as a maybe that has to prove it has been maintained on time rather than on miles.

Your instinct about the low-mileage car is the right one, and it is the thing most buyers get backwards.

Engines and gearboxes wear with use. Almost everything else degrades with time and with sitting. Rubber perishes, seals dry out and leak, brake discs corrode, fluids degrade regardless of miles, batteries die, tyres go hard and crack. A car that has done four thousand miles a year has spent nine years doing all of that with none of the compensating benefit of being run regularly.

So the nine year old car is not lightly used, it is old, and it has probably had the maintenance schedule of a car that never seemed to need anything. That is where the cost is hiding.

Meanwhile ninety thousand miles on a four year old car is high but entirely ordinary for the engine, particularly if they were motorway miles.

What actually decides it is the service history, not either number. A stamped, on-schedule history with cambelt and fluids done at the right intervals tells you more than the odometer. And on the low-mileage car, look for time-based servicing rather than mileage-based - many people with low mileage skip services because they have not hit the miles, which is precisely the wrong reading of the schedule.

The point about the kind of miles is worth expanding, because it does a lot of work here.

Motorway miles are gentle. Constant speed, engine at temperature, few cold starts, minimal braking, no stop-start clutch work. A car with ninety thousand of those can be in better mechanical condition than one with thirty thousand short urban trips.

Short cold trips are the harshest use there is. The engine never fully warms, oil dilutes with fuel, condensation accumulates in the exhaust and the oil, and diesel particulate filters in particular never regenerate properly.

So the questions to ask are not how many miles but what kind:

• What was it used for, and by whom
• How far was the daily journey
• Was it ever mostly at motorway speeds

A fleet or company car doing long trips is frequently a good buy despite the number, and this is the segment where the price drop is largest relative to the actual condition.

On the low-mileage car, the equivalent question is how it was stored and whether it was run. A car started once a fortnight for a short trip has had nine years of the worst possible use pattern in miniature. A car genuinely laid up and properly recommissioned is a different thing again.