PDF417

PDF417 is a stacked two-dimensional barcode defined by ISO/IEC 15438. On U.S. and Canadian driver’s licenses it stores a duplicate of the data printed on the front, encoded to the AAMVA DL/ID Card Design Standard. The name describes its structure: every codeword is four bars and four spaces across seventeen modules.

Full name Portable Data File 417
Symbology standard ISO/IEC 15438:2015
ID application standard AAMVA DL/ID Card Design Standard (v10, 2020)
Type Stacked linear barcode, read as two-dimensional
Symbol size 3 to 90 rows, 1 to 30 columns
Codeword structure 4 bars and 4 spaces across 17 modules, base 929
Maximum capacity 925 data codewords: roughly 1,033 bytes, 1,726 text characters, or 2,528 digits
Error correction Reed–Solomon, levels 0–8, using 2 to 512 correction codewords
Encryption None — contents are plaintext to any reader

How it works

A PDF417 symbol is a stack of rows, each one a self-contained linear barcode. Every row opens with a start pattern and a left row indicator, carries one to thirty data codewords, then closes with a right indicator and a stop pattern. Those indicators encode the symbol’s dimensions and error correction level, so a scanner reconstructs the whole from partial reads and need not see the rows in order.

Each codeword occupies seventeen modules split into four bars and four spaces — the source of the “417,” with PDF standing for Portable Data File. Reed–Solomon correction is why the format survives life in a wallet: at higher levels, scratches, creases, and a torn corner all reconstruct from the remaining codewords.

On a driver’s license the payload follows AAMVA’s structure rather than being free-form. It opens with a compliance indicator, the literal string ANSI, a six-digit Issuer Identification Number naming the jurisdiction, and a version number. A subfile designator — DL or ID — introduces the record, and every field inside carries a three-letter code: DCS for family name, DAC for first name, DBB for date of birth.

Why it matters for identity verification

The barcode’s value is redundancy. Because it duplicates what is printed on the front, any mismatch between the payload and the text extracted from the front of the document is a strong tampering signal. Alter a printed date of birth without re-encoding the barcode and the contradiction takes milliseconds to find.

That cross-check is the most useful thing PDF417 offers a verification flow. The barcode itself is not hard to produce; conformant encoders are freely available. What is hard is producing one that agrees with a physical card which also survives its own authenticity checks. Forged licenses fail so often because the two halves were built separately.

PDF417 also travels well. Designed for laser scanners, it decodes reliably from a phone photograph in ordinary indoor light — which is what makes remote onboarding practical at all. That cuts both ways. A photograph of someone else’s license scans exactly as well as one of your own, and the barcode has no way to object.

PDF417 vs QR code

  PDF417 QR code
Structure Stacked linear rows True matrix
Shape Wide rectangle, suits a card’s reverse Square
Standard ISO/IEC 15438 ISO/IEC 18004
Typical capacity Around 1,033 bytes Around 2,953 bytes
Orientation Needs approximate alignment Omnidirectional via finder patterns
Use on ID documents The AAMVA standard for North American licenses Rare on licenses; common on boarding passes and vaccination records

QR holds more data in less space and reads from any angle. PDF417 persists on licenses because the standard predates widespread QR adoption, and North American issuance infrastructure is built around it.

What it can’t do

It is not a security feature. This is the most common misreading of the format. PDF417 is a data container; the AAMVA standard specifies what goes in it, not who may write it. A barcode that scans and parses cleanly proves only that someone followed the specification.

It carries no cryptographic signature. Unlike an ePassport chip, which signs its contents against an issuing authority’s certificate, a PDF417 payload is plaintext with no way to trace it to the issuing DMV. There is nothing to validate against — only internal consistency to check.

It cannot detect a copy. The barcode reproduces perfectly through a photograph, a photocopy, or a screen. Nothing changes between the original card and an image of it, so it contributes nothing to whether a physical document was present — the question that matters most remotely.

It is frozen at printing. The encoded address, name, and restrictions describe the holder on the day of issue. A valid license may carry a barcode years out of date, and a mismatch against current records is at least as likely to mean the holder moved as that the card is forged.

Frequently asked questions

What data is stored in a driver’s license PDF417 barcode?

The same data printed on the front: name, date of birth, address, license number, issue and expiry dates, physical descriptors such as height and eye color, and any restrictions or endorsements. AAMVA defines which fields are mandatory per jurisdiction.

Is the PDF417 barcode on a driver’s license encrypted?

No. The contents are plaintext and any conformant scanner reads them. That is deliberate — the barcode exists so police, retailers, and agencies can read a license quickly without proprietary equipment.

Can a PDF417 barcode be faked?

Yes, trivially, in isolation. Encoding arbitrary data to AAMVA’s structure requires no special access. The difficulty is producing one consistent with a physical card that also passes the document’s own authenticity checks.

Why do U.S. licenses use PDF417 instead of QR codes?

AAMVA standardized on PDF417 before QR was widely deployed, and its rectangular shape suits the back of a card better than a square. Changing now would require every jurisdiction and every reader in the field to migrate at once.

Related reading

Discover Our Solutions

Exploring our solutions is just a click away. Try our products or have a chat with one of our experts to delve deeper into what we offer.

Report
Mapping the Rise of AI-Powered Identity Fraud

AI didn't just make fraud faster. It made it a system. We analyzed millions of identity interactions to map how identity attacks are evolving across regions, attack types, and sophistication levels — and what organizations need to rethink to keep pace.

See the Data