Used-Car Statistics 2026: Prices, Costs, Fraud and Records
US used-car prices were 2.3% lower in August 2026 than a year earlier, but maintenance and repair costs were 5.2% higher. NHTSA still estimates about 452,000 odometer-fraud cases a year. Price, ownership cost and vehicle-history risk must be measured separately.
Sections11
This is a data desk, not a claim that one number describes the whole used-car market. Price indexes, repair costs, fraud estimates, title records and mechanical inspections answer different questions. The tables below keep the period, population and limitation attached to each figure.
Downloads: source ledger CSV | monthly used-car CPI, 2020-2026
Used-car data desk
| If you need | Go to | Best source in this brief |
|---|---|---|
| Latest used-car price movement | Prices and inflation | BLS CPI, August 2026 |
| Purchase price versus running costs | Ownership-cost pressure | BLS CPI categories |
| Mileage-fraud scale | Odometer fraud | NHTSA incidence research |
| What VIN and title tools cover | Vehicle-record source atlas | NHTSA, NICB and NMVTIS |
| Dealer warranty disclosure | FTC Buyers Guide | 16 CFR Part 455 |
| Inspection-data definitions | Korean case study | VINLID's 1.55 million-report sample |
| Reusable rows and formulas | Downloads and methodology | Two versioned CSV files |
Latest US used-car statistics
| Finding | Period and scope | Status | What it means |
|---|---|---|---|
| Used cars and trucks: -2.3% | August 2025 to August 2026, US CPI-U | Observed, BLS | National price movement, not an average selling price |
| Used cars and trucks: +0.4% | July to August 2026, seasonally adjusted | Observed, BLS | One-month movement; subject to revision |
| Maintenance and repair: +5.2% | August 2025 to August 2026, US CPI-U | Observed, BLS | Running-cost pressure across motor vehicles |
| Motor-vehicle insurance: -5.1% | August 2025 to August 2026, US CPI-U | Observed, BLS | National index movement, not a driver quote |
| Gasoline: +27.4% | August 2025 to August 2026, US CPI-U | Observed, BLS | Fuel-price movement across all consumers |
| About 452,000 odometer-fraud cases | Annual US estimate | Observed, NHTSA | Scale of mileage fraud, not a listing-level probability |
| 92.49% of US earned insurance premium | Participating NICB members, August 2026 | Observed, NICB | Insurer participation scale, not vehicle-record completeness |
Used-car prices and inflation
The Bureau of Labor Statistics reported an August 2026 unadjusted used-cars-and-trucks CPI of 184.580, down 2.3% from August 2025. Its seasonally adjusted series was 180.965, up 0.4% from July after rounding in the official release.
Do not translate either index into an average used-car price. CPI measures how prices change for a defined basket. It does not say that the average vehicle sold for $18,096.50, and it does not separate dealer, private-party and auction transactions.
The downloadable monthly series preserves BLS's seasonally adjusted values from January 2020 through August 2026. Three transparent calculations put the latest index in context:
| Comparison | Formula | Result |
|---|---|---|
| August 2025 to August 2026 | (180.965 / 185.259 - 1) x 100 |
-2.32% |
| February 2022 peak to August 2026 | (180.965 / 216.876 - 1) x 100 |
-16.56% |
| January 2020 to August 2026 | (180.965 / 139.424 - 1) x 100 |
+29.79% |
The first result differs slightly from BLS's published -2.3% because our calculation uses the seasonally adjusted series and shows two decimal places. The official release's year-over-year table uses unadjusted indexes. Keep those series separate.
Ownership-cost pressure
A cheaper purchase index does not guarantee cheaper ownership. BLS reported these 12-month changes for August 2026:
| CPI category | 12-month change | One-month change | Scope warning |
|---|---|---|---|
| Used cars and trucks | -2.3% | +0.4% seasonally adjusted | Price movement, not transaction dollars |
| New vehicles | +0.6% | +0.3% seasonally adjusted | Comparison category, not used vehicles |
| Maintenance and repair | +5.2% | +1.1% | Not seasonally adjusted in this table |
| Motor-vehicle insurance | -5.1% | -0.8% seasonally adjusted | Not a quote for a driver, state or coverage level |
| Gasoline, all types | +27.4% | +3.9% seasonally adjusted | Volatile and not vehicle-specific |
These indexes cannot be added into one "cost of owning a used car" rate. Each category has a different weight and population. Their value is directional: acquisition prices, repairs, insurance and fuel can move in different directions at the same time.
For an individual purchase, compare the written out-the-door price, APR, term, amount financed, insurance quote, expected maintenance and a repair reserve. The FTC warns that a lower monthly payment can hide a longer term and higher total financing cost.
Odometer fraud
NHTSA defines odometer fraud as disconnecting, resetting or altering an odometer with intent to change the mileage shown. Its current page reports:
- More than 450,000 vehicles sold each year with false readings.
- Approximately 452,000 cases per year in the United States.
- A 3.47% chance during the first 11 years of a vehicle's life in its incidence research.
- More than 250 criminal convictions across more than 30 states.
- More than $2.8 million in criminal fines and $15 million in court-ordered restitution.
The enforcement totals have no start date on the page, so they are cumulative, not annual. The annual estimate and 11-year incidence must not be added together.
NHTSA recommends comparing the dashboard reading with the title, service and inspection records, oil-change stickers, tire age and visible wear. The used-car inspection checklist puts those checks in order.
Vehicle-record source atlas
There is no single source that proves a used car is safe or complete.
| Evidence source | What it can show | What a clean result cannot prove |
|---|---|---|
| NHTSA VIN recall search | Certain unrepaired safety recalls tied to a VIN | Completed recalls, some new recalls, most recalls over 15 years old, every small manufacturer or an international vehicle |
| NMVTIS consumer report | Title, brand, odometer, salvage and total-loss records supplied to the federal system | Present mechanical condition, every accident or every unreported event |
| NICB VINCheck | Participating insurers' unrecovered-theft, salvage and flood records | A complete title, repair, law-enforcement or accident history |
| FTC Buyers Guide | Dealer's written warranty or as-is position | Mechanical quality or future repair cost |
| Service and inspection records | Dated mileage and work recorded by the issuer | Work never recorded or records held elsewhere |
| Independent mechanic | Present condition within the agreed inspection scope | A complete lifetime history or a guarantee against failure |
NICB says its free tool checks a correct 17-digit VIN and allows five searches per IP address in 24 hours. Participating members represented 92.49% of US earned insurance premium as of August 2026. That percentage describes premium, not the share of vehicles with complete records.
NHTSA's VIN recall result has its own limits. A "0 unrepaired recalls" response does not display a recall already repaired, some recently announced campaigns, most campaigns older than 15 years, certain small manufacturers or international recalls.
FTC Buyers Guide
The FTC Used Car Rule, 16 CFR Part 455, has applied since 1985. It requires dealers to display a Buyers Guide stating whether the vehicle is sold as is or with a warranty. When a warranty applies, the guide states its duration, covered systems and the share of repair costs the dealer will pay.
The guide is a disclosure, not a safety certificate. Photograph it before negotiating and compare it with the final contract. The FTC also advises buyers to get the out-the-door price in writing and compare total financing cost rather than focusing only on the monthly payment.
Korean case study
VINLID reports analyzing 1,548,093 South Korean pre-sale inspection records, many involving vehicles listed for export. This is a proprietary case study, not an official census or a US benchmark.
| Measure in the VINLID sample | Share |
|---|---|
| Recorded accident history | 14.4% |
| Structural-member damage | 10.75% |
| At least one replaced body panel | 39.7% |
Korean inspection definitions can classify replacement of bolt-on parts such as a hood, fender or door as a simple exchange rather than an accident entry. Two calculations show why a yes-or-no accident field can capture less than a panel-level inspection:
39.7 - 14.4 = 25.3 percentage points.39.7 / 14.4 = 2.7569, or 2.76 times after rounding.
The definition lesson transfers; the rate does not. The source does not support a claim that 39.7% of US used cars have replacement panels.
What the available data cannot answer
No defensible current national dataset found in this review provides a single US used-car inspection failure rate, the share of independent inspections that stop a sale, or a combined probability of title, mileage, accident, flood and mechanical problems.
Those categories overlap and use different sampling frames. Adding them would double-count vehicles and create a statistic no source measured. Likewise, the public BTS used-vehicle table combines franchised-dealer, independent-dealer and casual sales and relies on industry-supplied source material; it is not used here as a real-time transaction count.
Downloads and methodology
The source ledger records each observed or derived figure with its period, geography, scope, source URL, access date, interpretation and limitation. The monthly price-index file contains one BLS observation per month.
Method:
- Prefer current US government sources for national price, safety, rule and system claims.
- Keep each source's population, period and definition beside its number.
- Label nonprofit and proprietary data instead of presenting it as government data.
- Publish the inputs and formula for every calculation.
- Refuse a combined risk score or unsupported market average.
Sources were accessed on September 14, 2026. Dynamic values are snapshots and may be revised. The original publisher remains authoritative for its current data.
Suggested citation:
Car Inspection Checklist, "Used-Car Statistics 2026: Prices, Costs, Fraud and Records," updated September 14, 2026, source page.
Primary sources
- BLS CPI Table 1, August 2026, modified September 11, 2026. Used for the published 12-month and monthly category changes.
- BLS used-cars-and-trucks CPI via FRED, updated September 11, 2026. Used for the seasonally adjusted monthly series and transparent calculations.
- NHTSA: Odometer Fraud, accessed September 14, 2026.
- NHTSA: Check for Recalls, accessed September 14, 2026.
- NICB: VINCheck, accessed September 14, 2026.
- US DOJ: NMVTIS, accessed September 14, 2026.
- FTC: Used Car Rule, accessed September 14, 2026.
- FTC: Financing or Leasing a Car, accessed September 14, 2026.
- VINLID: Korean Used-Car Accident Statistics, updated July 27, 2026 and accessed September 14, 2026.
- BTS: New and Used Passenger Car and Light Truck Sales and Leases, accessed September 14, 2026. Reviewed for scope and excluded from the headline snapshot.
Quick answers
Are used-car prices rising or falling in 2026?
BLS reported that its used-cars-and-trucks CPI was 2.3% lower in August 2026 than in August 2025, while the seasonally adjusted index rose 0.4% from July to August. CPI measures price movement, not the average selling price of a vehicle.
Are lower used-car prices making ownership cheaper?
Not across every cost. In August 2026, BLS reported used cars and trucks down 2.3% year over year, but motor-vehicle maintenance and repair up 5.2%. Insurance was down 5.1% and gasoline was up 27.4% over the same period. Those are national CPI movements, not quotes for one household.
How common is odometer fraud?
NHTSA estimates about 452,000 US cases per year and a 3.47% chance during a vehicle's first 11 years. These are population estimates, not a prediction for one listing.
Does a clean VIN check prove a used car is safe?
No. NICB says VINCheck searches participating insurers' theft, salvage and flood records and is not a comprehensive history report. NHTSA's VIN recall search also omits completed recalls, some newly announced recalls, most recalls over 15 years old and some small manufacturers.
Can I download and cite this data?
Yes. The source ledger records each figure's period, scope, source, access date and limitation. A separate monthly file contains the BLS used-cars-and-trucks CPI from January 2020 through August 2026. Cite the original publisher for a source statistic and this page for the compilation or calculations.