What Is PDF/A and When Do You Actually Need It?

PDF/A is a constrained version of PDF designed for one purpose: making sure a document looks exactly the same decades from now as it does today. It is defined by the ISO 19005 standard and is widely required by courts, government archives, libraries and regulated industries.

What PDF/A requires

A normal PDF can depend on things that may not exist in the future — a font installed on your computer, a colour interpreted by your specific monitor, a linked external file, or a script. PDF/A forbids exactly those dependencies. In practice a conforming file must:

embed every font it uses, so no substitution can occur;
declare its colour with an embedded output profile (an OutputIntent) so colour is device-independent;
carry XMP metadata identifying the conformance level;
avoid encryption, JavaScript, audio, video and external references.

The conformance levels

The number is the standard family and the letter is the strictness:

PDF/A-1b ("b" for basic) guarantees the visual appearance is preserved. This is the most widely accepted and the safest default.
PDF/A-1a adds structural tagging for accessibility, which is much harder to produce automatically.
PDF/A-2 and -3 allow newer features such as JPEG 2000 and embedded attachments.

If someone simply says "send it as PDF/A", they almost always mean PDF/A-1b.

When you actually need it

You need PDF/A when an institution tells you to — e-filing systems, tender portals, academic thesis repositories and records-retention policies often mandate it. For everyday sharing, a normal PDF is fine. Converting to PDF/A unnecessarily can even make a file larger.

How PDFelly handles it honestly

The PDF to PDF/A-1b tool embeds a real sRGB ICC profile as an OutputIntent, writes the required XMP identification, forces a PDF 1.4 structure, and runs a built-in self-check on its own output. Its "guaranteed" mode rasterises pages so there are no un-embedded fonts to violate the standard — the trade-off being that text becomes non-selectable. Because no in-browser tool can self-certify, we tell you plainly to confirm the result with a dedicated PDF/A validator before relying on it for compliance.

How PDF/A differs from a normal PDF in practice

Two files can look identical on screen yet only one is valid PDF/A. The difference is in the guarantees baked into the file: embedded fonts so nothing is ever substituted, an embedded colour profile so colours are unambiguous, identification metadata that declares the conformance level, and the absence of anything that depends on the outside world. A normal PDF may rely on a font installed on your computer; open it on a machine without that font and the text reflows or substitutes. PDF/A makes that impossible.

The trade-off you cannot avoid in the browser

Strict PDF/A demands every glyph be embedded. If a source PDF left fonts out, a browser cannot invent them. The honest workaround is to rasterise each page to an image, which has no fonts to embed and therefore always conforms — at the cost of selectable text. The alternative, keeping the text, only validates when the source already embedded its fonts and avoided transparency. Knowing this up front explains why "guaranteed" mode produces image-based pages.

Validating before you rely on it

Because the stakes for PDF/A are usually compliance-related, never assume — verify. A dedicated PDF/A validator is the recognised authority and will tell you exactly which clause, if any, a file fails. PDFelly runs an internal self-check on its own output to catch obvious problems and block a broken download, but a passing self-check is a sanity test, not a certificate.

Frequently asked questions

Is PDF/A-1b enough, or do I need PDF/A-1a?

PDF/A-1b preserves visual appearance and is what almost every requirement means by 'PDF/A'. PDF/A-1a adds accessibility tagging and is only needed when specifically requested.

Why is my PDF/A file larger than the original?

Embedding fonts and a colour profile, or rasterising pages, adds data. PDF/A optimises for permanence, not size.

Can I edit a PDF/A file later?

You can, but editing may break conformance. Re-validate (and if needed re-convert) after any change.

Does PDFelly certify PDF/A compliance?

No tool that runs in a browser can self-certify. PDFelly builds to the ISO 19005-1 spec and self-checks; confirm with a dedicated validator for regulated use.

Who actually requires PDF/A

It helps to know the institutions that drive most PDF/A requirements, because they explain when you genuinely need it. Court e-filing systems frequently mandate PDF/A so that filings remain readable for the full retention period of a case. Government tender and procurement portals often require it for submitted bids. University repositories commonly insist on PDF/A for theses and dissertations that must survive for decades. Corporate and public-sector records-retention policies specify it for documents that must be archived rather than merely stored. Libraries and national archives standardise on it for the same reason. Outside these contexts, everyday documents almost never need PDF/A, and converting unnecessarily can make files larger and, in image mode, strip their selectable text. The practical rule is simple: convert to PDF/A when a system or policy tells you to, choose PDF/A-1b unless told otherwise, and validate the output before submission so a rejected filing does not cost you a deadline.

Related guides

Try it now: Convert to PDF/A-1b — free, private, runs entirely in your browser. No upload, no account.