PDF417 Barcodes on North American Driver Licenses

If you verify identities in the United States or Canada, you will eventually deal with the dense black rectangle on the back of nearly every driver license. That is a PDF417 barcode, and it holds a structured, machine-readable copy of the cardholder data. Reading it correctly is one of the fastest ways to capture accurate identity fields and to catch documents that do not add up.

This post explains what PDF417 is, what it contains, and how to use it as part of a verification workflow.

What PDF417 Is

PDF417 is a two-dimensional stacked barcode format. Unlike a simple 1D barcode that stores a single number, PDF417 packs several hundred to a few thousand characters into rows and columns of small bars. On North American licenses it carries the AAMVA (American Association of Motor Vehicle Administrators) data standard.

The AAMVA standard defines a fixed set of data elements, each with a three-letter identifier. That structure is what makes the barcode so useful: instead of guessing where a field sits on the front of the card, you read named values directly.

Common AAMVA fields

• DAC / DAD / DCS — first, middle, and last name
• DBB — date of birth
• DBA — license expiration date
• DBD — issue date
• DAG / DAI / DAJ / DAK — street, city, state, postal code
• DAQ — license or ID number
• DCG — issuing country
• DCA / DCB — vehicle class and restrictions

There are more, and not every jurisdiction populates every field, but this core set covers most verification needs.

Why Read the Barcode at All?

You can OCR the printed text on the front of the license, so why bother with the barcode? Two reasons.

First, accuracy. Printed fonts, glare, and worn cards make front-side OCR harder. The barcode is a clean digital payload — when it scans, you get exact characters, not a best-guess interpretation of pixels.

Second, cross-checking. The most valuable thing about the barcode is not the data itself but comparing it to the front of the card. A genuine license should have matching name, date of birth, and license number on both the printed front and the encoded back.

Using PDF417 to Detect Fraud

A surprising number of counterfeit and altered licenses fail on the barcode. Common patterns:

• Mismatched front and back. A fraudster edits the printed name or date of birth on the front but leaves the original barcode untouched — or vice versa. A field-by-field comparison exposes this instantly.
• Missing or unreadable barcode. A card that should have a PDF417 but does not, or one that scans to garbage, warrants closer inspection.
• Malformed AAMVA structure. Fake barcodes often omit required subfile headers or use invalid data element codes.
• Impossible values. An expiration date before the issue date, or a date of birth that makes the holder implausibly young or old for the license class.

None of these checks is conclusive on its own, but together they raise or lower a confidence score quickly and cheaply.

Practical Parsing Notes

A few things trip up teams building their own parsers:

Layout varies by jurisdiction

While the AAMVA data elements are standardized, image quality, barcode density, and which optional fields are populated differ across the 50+ US and Canadian jurisdictions. Build for variation rather than assuming a single canonical layout.

Names are not always clean

Suffixes, hyphenated surnames, and truncated fields appear regularly. Normalize before you compare barcode names to front-side OCR, or you will generate false mismatches.

Dates come in different formats

AAMVA uses MMDDYYYY in most US jurisdictions but YYYYMMDD in others. Parse based on the issuing jurisdiction and standard version, not a fixed assumption.

The barcode proves data, not authenticity

Reading a valid barcode confirms the encoded data is well-formed and consistent. It does not confirm the physical card is genuine. For that, pair barcode reading with document authentication — checking security features, fonts, layout, and known template characteristics — and with a biometric face match against the portrait.

Where PDF417 Fits in a Full Workflow

Barcode reading is one signal in a layered check. A robust North American ID flow typically includes:

1. Document OCR of the front to extract printed fields.
2. PDF417 decode of the back to extract AAMVA data.
3. Cross-comparison of front OCR against barcode data.
4. Document authentication to assess forgery and tampering.
5. Biometric face match and liveness to tie the document to the person presenting it.

Each layer catches problems the others miss. Barcode reading is fast and inexpensive, which makes it an efficient early filter before more intensive checks.

Summary

The PDF417 barcode on North American driver licenses is a compact, standardized source of identity data. Read it to improve accuracy over front-side OCR alone, and — more importantly — to cross-check the front against the back for tampering. Just remember that a clean barcode confirms consistency, not authenticity. Combine barcode reading with document authentication and biometric verification for a complete picture, and normalize names and dates carefully to avoid false rejections along the way.