Taking the ratio between the top and bottom decile of price inside each model year — how far apart the cheapest and dearest examples of the same car sit.
| Model year | Price spread | Item |
|---|---|---|
| 2025 | 1.68x | Tight agreement |
| 2023 | 1.95x | |
| 2015 | 2.55x | |
| 2010 | 3.09x | |
| 2005 | 4.22x | |
| 1999 | 5.56x | No agreement at all |
Almost monotonic across 28 years
Each year older widens the gap, with only tiny reversals. A 2025 car's cheap and expensive examples are 68% apart. A 1999 car's are 456% apart.
Why old cars stop having a price
Condition diverges without limit. Two 2025 cars are both nearly new. Two 1999 cars might be a garaged 60,000-mile example and a rusted 250,000-mile one. Time doesn't average vehicles out — it spreads them.
Configuration matters more as value falls. On a new car, options are a modest share of price. On a $6,000 car, a desirable engine or drivetrain can double it.
Title status has room to move. The gap between a clean and a branded old car is proportionally enormous.
Sellers guess. With few comparables, some listings are simply wrong in both directions — and 20% of valuation requests hit a model year with no matching listings at all.
The trap this creates
A valuation on a recent car is genuinely informative — the market has converged and the number means something.
A valuation on a twenty-year-old car is a midpoint between two populations that barely overlap. Quote it in a negotiation and you're citing a number that describes almost none of the actual cars for sale.
Worse, thin data makes the range look tighter, not wider: with four listings there's no room for an outlier, so the numbers behind the least evidence look the most confident.
How to price an old car properly
Decide which end of the distribution you're in. Not what a 2005 Accord is worth, but what a well-maintained one-owner 2005 Accord with full records is worth — those are different questions with different answers.
Use configuration and condition as the argument. On an old car these justify a position at the top or bottom of a 4x range, which is worth far more than haggling.
Records are the main lever. Documented maintenance is what separates the top decile from the bottom, and it's the one thing a photograph can't fake.
Verify the history. On a spread that wide, a branded title or mileage discrepancy explains a lot of cheap listings. A VIN check shows which ones. We looked at how far this goes on individual models in the same 2001 F-250 selling for $2,300 and $38,000.
Frequently asked questions
How much do prices vary for the same used car?
It depends on age. Within a 2025 model year the top and bottom deciles are 1.68x apart; for a 1999 model year they're 5.56x apart. The spread widens almost monotonically across 28 model years.
Why do older cars have such variable prices?
Condition diverges without limit over time, configuration becomes a larger share of a smaller value, title status has proportionally more room to move, and with few comparables some sellers simply guess.
Can I trust a valuation on an old car?
Less than on a recent one. It's a midpoint between populations that barely overlap, so it describes almost none of the actual cars for sale — and thin data makes the range look tighter rather than wider.
How should I price a twenty-year-old car?
Decide which end of the distribution it belongs to. Configuration, condition and documented maintenance justify a position at the top or bottom of a fourfold range, which matters far more than negotiating around an average.
Denis Kataev is a serial solopreneur and the founder of ForCar. With 15+ years in software engineering and 10 years in SEO, he builds data-driven products end to end — backed by a sharp eye for design. At ForCar he mines proprietary vehicle datasets, turning raw numbers into buying advice you can actually trust.
Keep reading
All articles →
2,964 Complaints Come From Leased Cars. They Are Filed Earlier Than Anyone Else's.
Ownership changes what people report. Somebody who will hand the car back in two years has no resale value to protect and no repair bill to face, and it shows in the record.
62,142 Complaints State the Exact Mileage. 19,999 Mention the Warranty in the Same Breath.
Nobody asks for the mileage. A sixth of complaints supply it anyway, usually to the exact thousand, because the number is doing a job in the argument the owner is making.
1,372 Owners Filed a Complaint and Traded the Car In. 1,027 Said They Would Never Buy the Brand Again.
There is a moment when a fault stops being a repair problem and becomes a resale problem. A few thousand complaints record exactly that moment, and what it cost the person who reached it.
Mark ForCar as a preferred source and our data shows up first in your Google results. One click, no account.