Our Rule
If a page says "we tested", we ran the test, kept the numbers, and the page quotes those numbers. If we could not test something, the page says so. This page explains how that works, so you can judge our articles for yourself.
Where the Data Comes From
- Vehicle facts (make, model, plant, engine) come from the public NHTSA vPIC decoder. We query it slowly, about one request per second, and quote the dates we asked.
- The check digit follows the calculation set out in U.S. rules for VINs (49 CFR Part 565) and the ISO 3779 standard.
- Barcodes and QR codes are drawn with open-source libraries (JsBarcode and qrcodejs) and read back with open-source readers (ZXing and jsQR).
How We Run a Test
- Write down what we want to find out, before running anything.
- Build the test cases. When we need VINs we make samples with made-up serial numbers and a calculated check digit. Real vehicles' VINs are never used.
- Run every case through the real tool or the real data source, not a copy of it.
- Check the answers a second way when we can, for example with a separately written check-digit routine.
- Write the article from the recorded results. If the counts in the text do not match the log, the article does not go live.
Where it helps, we publish the raw material. Our 100-VIN test file has every case and result.
What We Have Tested So Far
| Question | How we answered it |
|---|---|
| Which VINs pass validation? | 100 test VINs in 8 groups through our validator, checked against a second routine |
| Does a QR code or a barcode survive damage? | Simulated scratches, smears and fading, read back with ZXing and jsQR |
| How small can a code be? | Every size and print-quality combination decoded in software |
| Code 39 or Code 128? | Size, tilt, damage and fade tests on 20 VINs |
| Where was my Hyundai built? | Live NHTSA decodes of sample VINs for each model |
| Where was my Kia built? | Live NHTSA decodes, plus a check of the maker list |
What We Cannot Tell You
- A VIN that passes our validator is internally consistent. That does not prove the car exists, who owns it, or that it has a clean history.
- Our code tests use software readers and simulated damage. They are not a phone or a scanner gun, so print a sample before you print a batch.
- NHTSA's data is strongest for vehicles sold in North America. For other markets a decode can come back blank, and we say so on the page.
Corrections
We fix mistakes in the open. If you find one, write to [email protected] and we will check it. Here is what we have changed:
| Date | What was wrong | What we did |
|---|---|---|
| 14 Nov 2026 | The VIN validator and decoder cut any text longer than 17 characters down to 17, so an 18-character VIN could be reported as valid | Raised the input limit so long text is rejected with its real length |
| 10 Nov 2026 | Our Hyundai article said NHTSA never names a plant city for Alabama-built cars. It does for some, such as Montgomery | Corrected the article and added the cases |
Who Does This
The site is run by Waqas Ahmad. Questions about a method are welcome through the contact page.