An old car's repair limit is what it sells for

In brief

• At the cheap end the median odometer reading of a car sold is 306,000 km and the median model year 2005.
• Price does not keep falling with age: the median price of a 2003 car is 1,490 € and of a 1996 car 1,200 €.
• The badge does not move the line. Mercedes-Benz E only reaches this price at 490,000 km.

The divisor is the price the car changes hands for

A repair decision is a division. The numerator is the garage's estimate, the divisor the car's current value. The estimate takes a couple of minutes on the phone, but nobody hands you the divisor, and that is where the decision falls apart. The same bill is an annoyance in a car worth ten thousand and the whole car in one worth a thousand. Without a divisor, rules of thumb are no help, because every one of them divides the repair cost by a value nobody has looked up.

Value here means one thing: the price somebody else pays for the car this week. It is not an insurer's figure, nor the sum the owner would consider fair. The cheap end sees plenty of trade: these cars account for 21.9% of all used-car sales in Finland, and 653 different models change hands among them. With that many sales, a comparison for your own car is almost always there, and it comes from the same week rather than from last year.

The divisor can be measured, and measured more precisely than the repair. For cars under 3,000 €, the median sale price is 1,599 €. Half of all sales fall below that and a quarter below 1,000 €. These are prices that were actually paid, not estimates.

Another way to see the divisor is to look at the market as a whole. Across all used-car sales the median model year is 2014 and the median odometer reading 187,000 km, so the typical car sold is nine years younger and more than a hundred thousand kilometres less driven than a car at the cheap end. The same repair means a different thing in those two cars even if the bill is identical to the euro, which is why the general sense of when repairing is sensible does not carry over to an old car.

The gap between those two figures is where the whole question starts. An old car drives the same roads, wears out the same brake pads and needs the same exhaust as one ten years newer, but it is worth a fraction as much. A garage's hourly rate does not look at the value of the car on the lift, and the price of a part does not bend to it either. So the ratio of repair to value turns against the owner in an old car even when the work itself is exactly the same as in a newer one. The owner of an old car is doing a different sum from the owner of a newer one, even if the bill looks the same and the garage is the same. The difference sits entirely in the divisor, and that is where the five most common mistakes bite.

The purchase price does not set the repair threshold

The first mistake is made before the garage is called back. A certain sum was once paid for the car, and that sum lodges in the head as the divisor. A price paid three years ago was the right figure on that day, but it described a car with three years fewer kilometres and one roadworthiness inspection (katsastus) less behind it. The older the car, the further the purchase price sits from what it would fetch today.

Model year explains the price only at the top of the bracket. At the cheap end the median price of a 2014 car is 2,590 € and of a 2006 car 1,690 €, so eight years cost less than a thousand euros. After that the curve flattens and finally turns: the median price of a 1998 car is 1,200 €, and below that the price climbs again. The year stops explaining the price at exactly the point where the repair question usually comes up for the first time.

At the top of the bracket the sum still works differently. The median price of a 2012 car is 2,300 € and of one a year older 2,200 €, so a year still costs something there, and a repair can lift the value enough to show up in a sale. The lower you go, the less of the price sits in the car itself and the more of it in the fact that the car runs and has passed its inspection. This is where the nature of the decision changes: at the top you repair property, at the bottom you repair mobility.

The upturn in the curve is not a measurement error. A thirty-year-old car is no longer bought as mere transport but because it is that particular car, and the price an enthusiast pays is formed on different grounds from a daily driver's. The median price of a 1991 car is 1,900 €, more than for a car fifteen years newer. The practical consequence runs both ways: age does not tell you the value, so a repair decision cannot rest on the year.

The same line shows in who buys these cars to sell them on. At the cheap end dealers account for 33.2% of sales of 2012 cars and 3.8% of sales of 1996 cars. The trade stops buying at roughly the point where the price curve flattens, and the rest of the bracket changes hands from one person to another. That is the single best marker of where a car's value stops being a calculated quantity and becomes an agreement between two people.

The purchase price is a poor divisor for two reasons. It is a figure from a past day, and it describes a different car from the one now on the lift. The only usable divisor is the price similar cars change hands for this week.

At the cheap end the badge buys no extra life

The second mistake is trusting the badge. The owner of a premium car reasons that it was built to last longer, so it is worth repairing for longer. The reasoning is not baseless, but at the cheap end it leads to the wrong answer. The badge does not bring the car to this price more slowly; it brings it by a different route, and the route is kilometres.

Mercedes-Benz E changes hands in this bracket at 490,000 km, while the median for the bracket is 306,000 km. It is the largest gap in the bracket. The same shows in other big models: Volvo V70 reaches this price at 464,686 km and Mercedes-Benz C at 391,000 km, while Peugeot 206 gets here at 194,500 km. The question is not which car is better, but how much of the car has already been used up before it costs this little.

The same logic applies to body type. At the cheap end the median odometer reading for an estate is 350,000 km and for a hatchback 250,000 km, even though the median prices are almost the same: 1,590 € and 1,550 €. A hundred thousand kilometres costs next to nothing at this end, because the price has already bottomed out. For a buyer that gap is an opportunity. For an owner weighing up a repair it is a reminder that the divisor stays the same even when the car is bigger.

In a repair decision that gap decides everything. With close to half a million kilometres on the clock, one repaired item is not the last repaired item, because everything else has covered the same distance. A part is still priced by model rather than by the car's present value, so an expensive car stays expensive to repair long after it has become cheap to buy. That combination is what makes a big premium car the one where the line is crossed first.

The badge is not useless information, then, but it works the other way round from the assumption. In this bracket an expensive badge means the divisor is the same as everyone else's and the numerator larger. The odometer says the same thing more directly, and it is there in the listing without anyone needing an opinion about the marque.

A diesel reaches this price at 395,000 km

The third mistake is the promise of the fuel. A diesel engine is said to last longer, and from that follows the conclusion that a diesel is worth repairing for longer than a petrol car. The claim about durability often holds, but in a repair decision it works back to front: precisely because a diesel lasts, it has covered more ground before it drops to this price.

At the cheap end the median odometer reading for a diesel is 395,000 km and for a petrol car 275,000 km. The price gap is small: the median price of a diesel is 1,700 € and of a petrol car 1,500 €. The durability has already been used up and paid for, and what is left of it is not the certainty a repair decision would rest on. The same engine that lasted three hundred thousand kilometres is now a four-hundred-thousand-kilometre engine.

The gearbox tells the same story on a smaller scale. At the cheap end the median odometer reading for an automatic is 350,000 km and for a manual 298,000 km, and the median prices are 1,800 € and 1,500 €. An automatic is both dearer and more heavily driven at this end, and that is exactly the combination in which one large repair collapses the sum fastest. In a manual car the same decision is made with smaller figures on both sides of the line.

The share of petrol cars says the same thing from the other direction. Petrol accounts for 68.4% of sales at the cheap end, against 44.4% across the whole market. A petrol car drops to this price older but less driven: its median model year is 2004 against the diesel's 2007. For a diesel owner that means the same-money alternative is often a petrol car a couple of years older with more than a hundred thousand fewer kilometres on the clock.

Two questions follow from this, in order. First: how much is left in the car, that is, where the odometer is heading over the next year. Then: what the same money would buy instead, and at what reading. For a diesel owner the answer to the second is usually a petrol car, because there are more than twice as many of them on offer at this end. The comparison is worth making across the fuel line even if your own car has always been a diesel.

In practice this changes what the estimate is measured against. A diesel repair bill is not best compared with a petrol repair bill but with what is left in the car. After four hundred thousand kilometres there is less left than after three hundred thousand, and for a repair decision that is the most important difference between the two.

The inspection follows the repair

The fourth mistake is timing. The repair is ordered for one fault, and the next roadworthiness inspection asks about everything else. According to the Finnish Transport and Communications Agency (Traficom), a passenger car is inspected for the first time 4 years after it enters use and every 2 years after that, but once more than 10 years have passed, the inspection comes round every year. In a car at the cheap end the next inspection is therefore always close.

The failure rate follows age closely. Traficom reports that in 2024, 29% of passenger-car roadworthiness inspections ended in failure. Among four-year-old cars 7% failed, but among 20-year-olds it was already 43%. In this bracket the median model year of a car sold is 2005, exactly the point on the curve where failure affects more than one car in three.

The inspection has a second role in a repair decision: it puts a date on it. What would otherwise be a matter for consideration until next spring has to be settled at once after a failure, and in a hurry there is less time to look at alternatives. Under Traficom's inspection interval an old car comes up every year, so that date lands twice as often as it does for the owner of a newer car. In practice it means the decision need not come as a surprise: its timing is known a year ahead.

In the oldest cars the figure turns back. According to Traficom, 38% of passenger cars at least 25 years old failed, fewer than among 20-year-olds. This is the same phenomenon as the upturn in the price curve: a car that has survived a quarter of a century in road use is a car that has been looked after, and it is not the same population as a twenty-year-old daily driver. The median car at the cheap end is firmly the latter kind, not the former.

One more question follows alongside the estimate: what the inspection has in store. One repaired fault will not get the car through if the reason for failure lies elsewhere, and two consecutive repairs within the same year are a different sum from one. The notes from the previous inspection are the best forecast available here, because they do not usually put themselves right. If the repair is done just before the inspection without anything else being checked, the decision is made on half the information.

Fixing a car for sale does not lift it above the ceiling

The fifth mistake is made when the car is already on its way out. It is taken to the garage to be made saleable, so it will fetch more, and the repair is justified by the sale to come. But the bracket has a ceiling, and repairs do not move it: three quarters of sales in this group stay below 2,280 €. The ceiling is not a property of one car but of what buyers pay for this age group.

At this end the buyer is usually another private individual. Private sales make up 82.3% of trade at the cheap end, while across the used-car market as a whole dealers account for 63.5%. A far smaller share of these cars passes through a dealer than of newer ones, and the gap widens the bigger and more heavily driven the car: for Mercedes-Benz E in this bracket the dealer share is 8.2%.

There is a price difference between dealer and private sales, but it is not a repair premium. The median price in dealer sales is 1,999 € and in private sales 1,500 €, and at the same time the median model year of a dealer-sold car is 2006 against 2004 for a private one. The gap follows the car rather than the seller, and repairs do not shift it. When the repair is done for the sale, the money goes into a car that stays under the ceiling anyway.

In practice this changes where making a car saleable is worth aiming. The buyer at the cheap end is usually a private individual who does part of the work themselves and prices a known fault into their own calculation. A car whose shortcomings have been stated and whose inspection is valid will change hands. A car that has just had a big repair changes hands at roughly the same price, but the seller paid for the repair. The difference between the two is one whole repair bill, and it shows in the price in neither direction.

This does not mean a car is not worth putting right before a sale. It means the reason for putting it right is getting the sale done, not raising the price. What the car will fetch was settled before the garage was ever called.

A big repair just before a sale does not raise the price. It moves money from the seller to the buyer.

Three figures before the repair is ordered

The sum can be done before the garage is called back. It needs three figures, and all three can be had the same day without moving the car anywhere. The first two concern your own car and the third what would replace it.

• What the same model, the same year and roughly the same reading sells for now. This is the divisor, and the price the car was once bought for is no substitute.
• When the car is next due for inspection and what was noted last time. A repair that does not get the car through the inspection is half a repair.
• What the same money would buy instead. The alternative is not a new car but another car of this age group, and its odometer is worth checking too.

The first figure is the easiest to find and the easiest to get wrong. Only a car whose model year and odometer reading are in the same range as your own works as a comparison, because those two are what separate cars at this end. One search result is not enough, since any single reason can lie behind one car's price. Ten similar cars are enough, and you take the middle one, not the cheapest and not the dearest. The same applies in the other direction when you look at what could replace it.

The third figure is the one most often left unchecked, and it is precisely the one that makes the comparison honest. The replacement car comes from the same pool, and its odometer shows roughly the same reading as your own. Swapping does not remove the risk; it moves it into a car whose history you do not know.

Figures do not settle everything, nor are they meant to. The car may be exactly the car you do not want to part with, or there may simply be no time to find another before the present one has to run again. The sum does not tell you what to do. It tells you what the decision costs, and that is the whole difference: knowing the price, you make the choice yourself rather than on the strength of a garage estimate. Finding three figures takes about as long as one phone call.

With the three figures side by side, the decision comes without a rule of thumb. If the estimate is larger than the price similar cars change hands for, the money goes into something whose value will not rise at all. If the estimate is clearly smaller and the inspection has just been passed, the repaired car is the cheapest way to carry on for another year, because the alternative costs at least as much. The grey area in between is settled by which repair of the year this one is.

Cross the line and the replacement is a car of the same age

Crossing the line does not mean the car should go to the scrapyard the same day. It means the money put into the repair will not come back in the sale price, and that the decision is better made with that knowledge than against it. Plenty of people knowingly drive a car that is not worth repairing, and that is a different thing from believing it is worth repairing.

One weighty thing argues against the claim. The replacement car comes from the same pool: its median odometer reading is 306,000 km and its median sale price 1,599 €, so a swap gets you another old car. A known fault in a known car may well be cheaper than an unknown fault in an unknown one, which is why the ratio does not settle any single case. It only tells you which side of the line you are on.

Below the line the sum forbids nothing. When the divisor is a thousand euros and the numerator hundreds, repairing is still the cheapest way to get about, because no alternative comes in under that: swapping costs too, and the car you swap into is much the same. The line concerns the one bill that approaches the price of the whole car, and in an old car such a bill arrives sooner or later. The decision is easier made in advance, by naming the sum you will not go past with this car.

The difference lies in how many times the bill repeats. A repair decision is not made once but over and over, and each time it is measured against the same divisor. When the same car needs a second and a third repair within one year, the divisor has not changed but the numerator has grown, and by the third time the total has passed the value of the car even though no single bill did.

The practical rule fits in one sentence: look first at what the car is worth, only then at what the repair costs. The line is met most often where cars under 3,000 € change hands, and what a step up costs is paid in kilometres as much as in euros, as the comparison between price brackets shows.

See what the cheap end has on offer right now