AAMVA DLDV state coverage: What “all 50 states” really means

Somewhere in the middle of most driver's license verification evaluations, a compliance officer asks the vendor a simple question: does this work in every state?

The vendor says yes. Both of them believe the conversation is over. It is not, because "works" means two different things, and the gap between them is where onboarding programs quietly develop a blind spot the size of California.

One vendor's "all 50 states" means: we can photograph, read, and extract data from any state's license. Another vendor's "43 states" means: we can confirm the license data actually exists in the issuing DMV's records. The first claim is about reading the card. The second is about verifying the record behind it. A program that needs the second and buys the first passes every demo and fails its first serious fraud wave.

At Signzy we sit through these evaluations from the vendor side of the table, and the state-coverage question is the single most common place where a buyer's assumption and a vendor's claim quietly diverge. This piece is the map: what each coverage number means, what the published numbers actually are, and the questions that force every vendor, us included, to answer precisely.

Related Solution

Driver's license verification APIDocument OCRIdentity Verification

Driver's license extraction coverage vs DMV record-check coverage

Driver's license verification runs on four verification methods, but for coverage purposes they collapse into two categories.

Extraction coverage is about the document. OCR, PDF417 barcode parsing, and template checks read what the license says. Because this only requires the card in front of the camera, it genuinely works for all 50 states, and for international documents too. When a vendor says "all 50 states," this is almost always the claim being made.

Record-check coverage is about the source. The AAMVA Driver's License Data Verification (DLDV) service checks submitted license data against the issuing DMV's records in real time and returns match/no-match flags per field. This requires the state to participate, and not all of them do. DLDV participation reached roughly 43 states plus Washington, DC by 2025.

The difference in what they prove is not subtle:

An AI-generated fake license is the cleanest illustration. It photographs perfectly and its barcode parses cleanly, so extraction passes it. Its license number was invented, so a record check kills it. If your threat model includes fabricated documents, the record-check number is your real coverage, and it is not 50.

AAMVA DLDV state coverage by verification provider

Coverage claims are scattered across product pages, documentation, and press releases, and they are rarely presented side by side. Here is what the major US vendors state publicly about their own DMV/AAMVA record-check coverage, as of August 2026, each figure from the vendor's own published materials:

Three things stand out in that table.

First, the honest vendors disagree with each other by up to four states. That is not sloppiness. Participation genuinely varies by vendor contract, use case, and date, and some states impose consent or use-case restrictions that make a state "covered" for one deployment and not another.

Second, the most useful row is the one that names the exclusions. A vendor that tells you which states it cannot check has told you something you can plan around. A vendor that only says "all 50 states" has told you nothing you can verify.

Third, California appears twice: once as the most commonly named exclusion, and once as the state one vendor built a proprietary workaround for. That is not a coincidence.

Why California matters in DMV verification coverage

The states most often named outside DLDV coverage are not small ones:

Add those up and the record-check gap covers roughly a quarter of America's licensed drivers, concentrated in the two states with the largest fintech, gig, and marketplace user bases. A verification program can be "43 out of 50 states" by count and miss a quarter of its applicants by volume.

Fraud operations understand this arithmetic. A fabricated license is far more likely to survive verification when it claims to be from a state where no record check is possible, because there the decision falls back entirely on how the image looks. If your fraud team has never segmented fake-document catches by claimed issuing state, that segmentation is worth running this week: a cluster of fakes claiming California or New York issuance is the signature of attackers who did their homework.

Six questions to ask about driver's license verification coverage

The point of this piece is not that any vendor is lying. Most coverage claims are true sentences about different things. The fix is to ask questions that cannot be answered with a blended claim. Put these six in your RFP or your next vendor call:

1. Split the claim. "In which states do you extract data, and in which states do you verify the record against the issuing DMV?" Two lists, not one.
2. Name the exclusions. "Which states can you not run a record check in today?" A precise vendor answers in one email.
3. Define the silent path. "When an applicant's license is from a non-participating state, what does your API return, and is the fallback flagged in the response?" The dangerous configuration is a flow that quietly downgrades to image-only checks without telling your risk engine it did so.
4. Ask what the record check verifies. Name, date of birth, license number, expiry? Field-level match flags or a single verdict?
5. Ask what it does not verify. A record check confirms the data exists, not that the submitter owns it. What pairs with it for possession: face match, liveness?
6. Date the answer. "As of which month is this coverage list current?" Participation changes; two states were added to DLDV in a single 2024 announcement. A list without a date is a list you cannot audit.

A vendor that answers all six precisely is showing you operational maturity. A vendor that returns to "we support all 50 states" has answered a question you did not ask.

How to handle states outside AAMVA DLDV coverage

Coverage is not a scoreboard where the biggest number wins. It is a risk-allocation decision. The practical posture that follows from the published numbers:

For applicants in the ~43 participating states, demand the record check and treat its match/no-match as a hard signal. For the exclusion states, know exactly which fallback runs, weight it accordingly in your risk engine, and compensate with the layers that do not depend on state participation: document forensics, face match with liveness, and identity-data correlation. That layered check is how our driver's license verification is built to run inside a complete onboarding flow, one call in the broader US API stack, with the field-level response documented in the API reference.

How to evaluate driver's license verification state coverage

Before your next vendor renewal or evaluation, open your current provider's response documentation and answer one question from it: for a license issued in California, does your verification verify the record or only read the card? If the documentation cannot answer that, send question 3 from the list above to your vendor today. The answer determines whether a quarter of your applicants are getting the verification you think you bought.