Lower mileage generally raises a collector car's price, but the effect is uneven, strongest at the low end of the odometer, and smaller than the differences between individual cars.
That is the honest summary of what the sold data shows. It is less tidy than a depreciation table, and more useful.
The car we measured
To look at mileage on its own, you need many sales of one car that otherwise varies little. The 1984-1989 Porsche 911 Carrera 3.2 coupe is a good candidate. It was built in large numbers, it sells often, and most examples are close in specification.
We took every Carrera 3.2 coupe recorded as sold on Bring a Trailer between 1 November 2023 and 26 October 2026, the most recent sale in the set as of 1 November 2026. That was 243 sales. We set aside 66 that were not stock coupes or could not be measured, which left 177. The method section below lists exactly what was removed.
What the numbers show
Grouped by recorded mileage, the 177 sales fell like this.
| Recorded mileage | Sales | Median | Middle half |
|---|---|---|---|
| Under 50,000 | 27 | $107,000 | $93,000–$136,500 |
| 50,000-99,999 | 81 | $85,000 | $71,956–$96,750 |
| 100,000-149,999 | 51 | $73,000 | $65,000–$86,500 |
| 150,000 and over | 18 | $70,500 | $60,000–$75,250 |
Across the whole set, the median was $82,500 and the middle half sold between $68,500 and $97,500.
The direction is clear. Each step up in mileage lowers the median, and the gap between the lowest-mileage group and the next is the largest: $22,000 between the medians.
Where the effect flattens
The difference between the 100,000-mile group and the 150,000-mile group is $2,500 at the median. Past a certain point, buyers seem to stop paying for each additional mile avoided and start paying for the car's condition and records instead.
Read that with care. The highest group has only 18 sales, which is too few to say much with confidence.
How noisy it is
The middle-half ranges overlap. A car in the 100,000-149,999 group at the top of its range sold for more than a car in the 50,000-99,999 group at the bottom of its range.
The overlap runs further than that. Four of the 51 cars with 100,000-149,999 miles sold for $107,000 or more, at or above the median of the lowest-mileage group. Four of the 27 cars under 50,000 miles sold for $73,000 or less, at or below the median of the 100,000-mile group.
Measured as a rank correlation, the relationship between mileage and price in this set is about -0.46. That is a real tendency, of moderate strength. It is not a rule you can apply to one car.
Why mileage is not the whole story
A table grouped by miles carries every other difference between the cars with it, so it is worth checking the obvious one.
The 1987-1989 cars have the G50 gearbox, which buyers tend to prefer. If the low-mileage group were mostly G50 cars, the gearbox could be doing the work. It is not: 70 percent of the under-50,000-mile cars are 1987-1989 models, against 69 percent of the 100,000-149,999 group, so the mix barely changes.
Looking at the 114 cars from 1987-1989 on their own, the pattern holds. The 19 under 50,000 miles had a median of $125,000, the 50 between 50,000 and 99,999 had $90,706, the 35 between 100,000 and 149,999 had $76,700, and the 10 over 150,000 had $72,706. Those groups are small, and the last one is very small.
What the table cannot separate is the kind of life behind the miles. Records, paint and ownership history are disclosed in listing text, not in a field we can count reliably. The cleanest way to say it: mileage is a strong signal of the kind of life a car has had, and the price follows the life as much as the number.
What this means for a single car
A low odometer reading is worth most when it is documented. A car with records that support the mileage belongs with the low-mileage sales. A car whose mileage is unknown or disputed does not, and we removed such cars from this set for that reason.
Once a car is well into six figures on the odometer, its condition, its maintenance history and its originality will usually move the price more than another twenty thousand miles would.
And any single-number adjustment, so many dollars per thousand miles, should be treated with suspicion. The sold data does not move in a straight line.
Method
Source query: completed auction results for Porsche 911, source model label "911 Carrera 3.2", body style coupe, source Bring a Trailer, status sold, sorted by date. Of 650 such records as of 1 November 2026, 243 ended between 1 November 2023 and 26 October 2026.
Removed before analysis, 66 in total: 17 Targas filed as coupes, 13 turbo-look (M491) cars, 12 backdates, 6 safari builds, 5 cars with larger non-original engines, 4 cars with total mileage disclosed as unknown, 3 project cars, 1 with no recorded mileage, 1 RUF, 1 slant-nose, 1 Club Sport, and 2 cars outside the 1984-1989 model years. Kilometre readings were converted to miles. Mileage is as recorded in the listing.
Medians and quartiles are calculated on sale prices as reported by the auction. The rank correlation is Spearman's, between mileage and price across the 177 sales. The 1987-1989 subset is the 114 of those 177 sales with model years 1987 to 1989.
This analysis covers one model on one auction site. It does not establish the mileage effect for other cars. The broader 911 market, across sources and model variants, is on the Porsche 911 1974-1989 page, with its rules in its method section. Other models are listed at /value/models.
Blog entries are general information, not financial, legal or purchasing advice. BidBetter figures are estimates, not appraisals. See the Disclaimer.