What a passport MRZ is, how its check digits work, what a selfie-to-document match confirms, and what neither check proves on its own.
If you run seller verification on a marketplace, you will hear two terms a lot: the MRZ and the selfie check. They sound technical, and vendors often describe them in ways that blur what they actually do. This post explains both in plain terms, so you can judge what a verification result is telling you and where its limits are.
The machine readable zone, or MRZ, is the block of text at the bottom of a passport's photo page. It looks like capital letters, numbers and a lot of < characters. It exists so that machines at borders can read the key details quickly and consistently.
Its format is set internationally by ICAO Document 9303, Machine Readable Travel Documents. Part 4 of that document covers machine readable passports, which ICAO calls the TD3 size, and Part 3 covers specifications shared by all machine readable travel documents (ICAO Doc 9303). ID cards use smaller formats, TD1 and TD2, with different line layouts.
A passport MRZ (the TD3 format) has two lines of 44 characters each. Between them, they carry:
Names are written in capitals with < used as a filler and separator, which is why a name like "Mary Ann Smith" appears as something like SMITH<<MARY<ANN.
Check digits are the clever part. Several fields in the MRZ, including the passport number, date of birth and expiry date, are followed by a single digit calculated from that field. There is also a composite check digit calculated across several fields together.
The calculation is simple:
< counts as zero.If someone mistypes or alters a single character in a field, the check digit will almost always stop matching.
This is how TrustVerify handles it. Our passport MRZ parser implements the ICAO 9303 TD3 check-digit calculation with exactly those weights and values. When a line's check digits fail, the parsed name and document number are withheld rather than passed on as an identity. A line that fails its own internal check is not treated as data about a person.
This is the part that is often oversold.
A valid MRZ proves that the text is internally consistent. The check digits fit the fields. It also gives you clean, structured data to compare against what the seller typed into your form: does the name match, does the date of birth match, has the passport expired?
A valid MRZ does not prove the passport is genuine. Check digits are public arithmetic. Anyone producing a forgery can compute them. It also does not prove that the person holding the passport is the person named in it.
So the MRZ is a strong consistency and data-quality check, and a useful tripwire for crude alterations. It is one input, not the whole answer.
The selfie step addresses the second gap: is the person submitting the document the person in the photo?
The seller takes a photo of themselves during the verification flow, and that image is compared with the portrait on the document. If the faces match closely enough, the check passes on that point. If not, it fails or goes to review. TrustVerify's KYC check works this way: the selfie is matched to the document photo. The selfie checks guide in our docs covers how the step fits into the flow, and the capture guide covers image quality.
A face match answers one question: does the face in the selfie agree with the photo on the document? It does not prove the document is genuine, and on its own it does not tell you who is behind the camera. Treat it as one signal among several.
If your risk assessment calls for more assurance on the selfie, for example because you see organised fraud using stolen identity documents, write that down as a separate requirement and ask any provider for specific evidence of what each of its selfie checks measures.
Each check closes a different gap:
| Check | Answers | Does not answer |
|---|---|---|
| MRZ check digits | Is the machine-readable data internally consistent? | Is the document genuine? |
| Data comparison | Does the document match what the seller told us? | Is the seller the document holder? |
| Expiry check | Is the document in date? | Anything about the holder |
| Selfie-to-document match | Does the person's face match the document photo? | Is the document genuine? |
| Sanctions and PEP screening | Does this person appear on a watchlist? | Is the identity real? |
The value is in the combination. A seller whose MRZ is consistent, whose details match the form, whose document is in date, whose selfie matches the photo and who is clear on AML screening is a much better bet than one who has passed any single step.
And for sellers trading as businesses, add a check on the company itself. Our post on KYC vs KYB vs KYS explains where each fits.
Comparing providers? Read our guide to choosing KYC software in the UK.
If you are designing identity checks for sellers, it helps to see exactly what each step returns. Read about TrustVerify identity verification.
KYC, KYB, AML screening and fraud prevention in one place, with published pricing.
KYC
Documents a Marketplace Can Accept for KYC, and How to Handle the Rest
Which ID documents marketplaces commonly accept for seller KYC, what to do with documents outside that list, and how to get clear images first time.
KYC
Proof of Address Checks: What They Prove and What They Don't
What a proof of address document actually shows, its limits for marketplace seller checks, and how to use it alongside stronger signals.
KYC
Selfie-to-Document Matching: What It Confirms and Where It Stops
What a selfie ID match tells a marketplace, what it cannot tell you, how to read the result and how to design around its limits.