Limited Time: Get your Carfax and Autocheck for $4.50 when you sign up! - instant delivery
CheapCarfaxAutocheck
Pricing Membership FAQ Support Blog
← All articles / Buying Guides
Buying Guides August 9, 2026 · 7 min read

How to Read an AutoCheck Score (What 85 Actually Means)

An 85 is not a test grade — it is a comparison against similar vehicles. How the AutoCheck score and its range actually work, and the one rule that keeps the number from misleading you.

CE
CheapCarfaxAutocheck Editorial Team
Vehicle History Research Team
How to Read an AutoCheck Score (What 85 Actually Means)

Every AutoCheck report leads with a number. Buyers see "85" and instinctively read it like a test grade — as if 85 means the car is 85% good. That is not what the score says, and misreading it is how people talk themselves into the wrong car, or out of the right one.

What the score actually is

The AutoCheck Score compresses a vehicle's recorded history into a single number on a scale that runs to roughly 100. But the number on its own is only half of the feature. The report pairs it with a similar-vehicle score range — the band that cars of the same class and age typically land in — and that range is what makes the number readable.

This is the part most buyers miss: the score is cohort-relative. A vehicle is scored against comparable vehicles — trucks against trucks of similar age, luxury sedans against luxury sedans — not against every car on the road. An 85 on a ten-year-old pickup and an 85 on a three-year-old commuter sedan are two different statements, because the cohorts behind them expect different histories.

The one-sentence version
Don't ask "is 85 good?" Ask "where does 85 sit inside this car's similar-vehicle range?" Above the range is cleaner than typical; below it, the records are dragging.
Get articles like this every week
Field guides, red flags, and the occasional discount.

How the range works — and what moves the score

Read the score as a position, not a grade. At or above the top of the similar-vehicle range means the car's paper trail is cleaner than what's typical for its kind. Inside the range is unremarkable — which, for a used car, is fine. Below the bottom of the range means something in the record is pulling it down, and your job shifts from admiring the number to finding the line items responsible.

The kinds of records that move the score are the ones you'd expect:

  • Reported accidents and damage entries, weighted by what was recorded about them
  • Title brands — salvage, rebuilt, flood, lemon-law buyback
  • Odometer problems: discrepancies, broken sequences, true-miles-unknown flags
  • Heavy-use classes — rental, fleet, commercial, taxi
  • Ownership churn: many owners in few years reads worse than one long stewardship

Just as important is what the score is not. It is not a mechanical inspection, not a prediction that the transmission will survive, and not a value estimate. The score reads paper, not pistons. A car maintained obsessively by an owner whose mechanic never reported to any database gains nothing; a garage queen that once ran through a rental fleet loses points it may not deserve. The score is a summary of records — exactly as good as the records, and no better.

A worked example makes the relative logic concrete. Take two family SUVs of the same model year carrying the same score, both sitting inside their similar-vehicle range. On paper they're twins. Open the records and they diverge: the first earned its deductions from one minor reported accident, followed by years of a single careful owner. The second has spotless damage history but spent its first three years in a rental fleet with the owner churn to match. Same number, different risks, different negotiations — the first car's question is "how good was the repair?", the second's is "how hard is a rental life on this drivetrain?" The score put both cars on your shortlist; only the records tell you what to ask about each one.

Why a high score with a bad detail still matters

Compression loses information. That's not a flaw in AutoCheck's method — it's the nature of turning a multi-page history into one number. A single severe record can coexist with a respectable score when the rest of the history is spotless, especially on an older vehicle where the cohort's expectations are already forgiving.

So use this rule: the score screens candidates; the records make decisions. Never let a good number overrule a bad line item. A salvage or flood brand, a structural-damage announcement from an auction lane, an odometer that reads lower than an earlier entry — those are walk-away findings no matter what number sits at the top of the page.

A score is a reason to keep reading or stop reading. It is never a reason to stop thinking.

The inverse case is worth knowing too. A mediocre score sometimes sits on an honest car: a five-owner history where every owner was short-term but legitimate, or a fleet sedan that was serviced relentlessly and driven gently on highways. When the number is low, read why it's low. If the deductions come from use-class and owner count rather than damage, you may be looking at the rare thing the used market actually offers — a discount without a defect.

Score vs CARFAX's no-score model

CARFAX made the opposite design choice: no single number, just the record timeline. Neither approach is wrong — they're built for different moments in a purchase. The AutoCheck Score is triage: when you're comparing six candidate cars on a Saturday morning, a number with a range attached tells you in seconds which histories deserve your afternoon. The CARFAX presentation is depth: when you're down to one finalist, the maintenance narrative matters more than any summary.

The two reports also differ in what they see, not just how they present it — AutoCheck's Experian backbone covers auction lanes CARFAX often misses, while CARFAX's service-shop network runs deeper on maintenance. We've broken that down record type by record type in our CARFAX vs AutoCheck comparison if you want the full picture.

One field note about how the score gets used against you: dealers quote it selectively. When a listing advertises the score, it's because the number flatters the car — and a quoted score with no report attached is an invitation to take the summary and skip the evidence. Treat an advertised score the same way you'd treat "one owner, garage kept": as a claim to verify, not a fact delivered. Ask for the full report, or pull it yourself; if the ad doesn't mention a score at all, that silence is worth the few dollars it costs to break.

A practical reading order

When an AutoCheck report lands in front of you, this order gets the most out of it in the least time:

  1. Score vs range — ten seconds. Above, inside, or below the similar-vehicle band?
  2. Title and brand section — the hard stops live here. Any brand means the score no longer matters.
  3. Odometer sequence — readings should only ever climb. A backwards step is a walk-away.
  4. Use class and owner count — rental or fleet history isn't fatal, but it should shape the price.
  5. Accident and auction entries — dates and descriptions, matched against what the seller told you.

Then decide whether the car stays on your list. If it does, that's the moment the report has already paid for itself — and at wholesale prices, that's a low bar. You can get an AutoCheck report for a few dollars instead of the $29.99 retail rate, which changes the economics of screening: at that price you can afford to run the report before the test drive, on every serious candidate, not just the finalist.

If you're still deciding whether AutoCheck is the right report for your situation, our cheap AutoCheck guide covers how the pricing and the report itself work end to end. And when there's a specific VIN in your hand, run a VIN search and see what's on file for that exact car.

The score is a genuinely useful instrument once you stop reading it as a verdict. It points; the records prove. Keep those jobs separate and the number will save you time without ever costing you money.

Run a report
Don't buy without checking the VIN
Carfax + AutoCheck from $4.50 · Instant · 60 seconds
Run a report →