Will This File Still Open in Twenty Years?
Most of the time, nobody worries whether a PDF will survive. You make it, you send it, it gets read, and that's the end of its life. But some documents have to last: a signed contract that must stay readable for the length of an agreement, a court filing, a medical record, a tax document, a government archive, a dissertation deposited in a university library. For those, "it opens today" isn't good enough. The question is whether it will open, and look identical, long after the software that made it is gone.
That's the problem PDF/A was created to solve. It's a version of PDF engineered for one goal: long-term preservation. A PDF/A file is a promise that the document will render the same way in fifty years as it does now — no missing fonts, no broken colors, no dependency on software or servers that may no longer exist.
What PDF/A Actually Is
PDF/A isn't a separate file format. It's an ISO-standardized subset of PDF (ISO 19005) — a regular PDF that follows a strict set of extra rules. Because it's still a PDF, a PDF/A file opens in any PDF reader you already use. The difference is entirely in what it's *allowed* to contain.
The guiding principle is self-containment: a PDF/A must carry everything needed to reproduce itself, and must avoid anything that could fail over time or depend on the outside world. Everything else follows from that one idea.
The Rules That Make It Last
Each PDF/A restriction exists to remove a specific long-term risk:
Every font must be fully embedded. A normal PDF can just *name* a font and trust the reader's computer to have it — which, as we cover in why your PDF looks different on another device, is exactly how layouts break. PDF/A forbids this. All fonts travel inside the file, so the text is guaranteed to render correctly forever.
Color must be device-independent. Colors are defined with embedded ICC color profiles rather than left to the display to interpret, so a red stays the same red on any screen or printer, now or later.
No encryption. PDF/A prohibits password protection and DRM. An archive has to stay openable — a document locked behind a password whose key is lost is a document destroyed. This creates a genuine tension: you cannot password-protect a file and keep it PDF/A-valid, which is worth knowing if you're used to protecting PDFs with encryption.
No external dependencies or active content. No JavaScript, no audio or video, no links to external files, nothing executable. Anything that could behave unpredictably — or simply stop working — as technology changes is banned.
Standardized metadata is required. Every PDF/A must include an XMP metadata packet that declares its conformance (which version and level). Interestingly, this is the opposite of the privacy-driven instinct to strip metadata we describe in PDF metadata and what's hidden in your documents — for archiving, standardized, self-describing metadata is a feature, not a leak.
The Versions
PDF/A has evolved across four generations, each built on a newer base version of PDF:
| Version | Year | Based on | What it added |
|---|---|---|---|
| PDF/A-1 | 2005 | PDF 1.4 | The original, strictest rules |
| PDF/A-2 | 2011 | PDF 1.7 | JPEG2000, transparency, layers, Unicode level |
| PDF/A-3 | 2012 | PDF 1.7 | Allows embedding any file type (e.g. a source spreadsheet or invoice XML) |
| PDF/A-4 | 2020 | PDF 2.0 | Modernized and simplified conformance |
A newer version isn't "better" — it's a different balance. Many archives still standardize on PDF/A-1 or A-2 precisely because they're the most conservative and widely supported. PDF/A-3's ability to embed the original source file inside the archival PDF is what powers hybrid formats like electronic invoices (a human-readable PDF with machine-readable data tucked inside).
Conformance Levels
Within a version, PDF/A defines conformance *levels* that trade off how much the file guarantees:
| Level | Guarantees | Use when |
|---|---|---|
| B (Basic) | Faithful visual reproduction | You need it to *look* right forever |
| U (Unicode) | Level B + text is searchable/extractable | You also need reliable search and copy |
| A (Accessible) | Level U + tagged structure for screen readers | You also need accessibility and reflow |
Level B is the most common and the minimum. Level A is the most demanding, because it requires a fully tagged document structure — the same tagging that makes a PDF accessible to screen readers. (PDF/A-4 reworked this scheme with variants for embedded files and engineering data, but the B/U/A idea captures the core trade-off.)
Who Needs It — and Who Doesn't
PDF/A is worth the extra constraints when a document has a *retention requirement* or long-term legal weight:
- Legal — many courts require e-filings in PDF/A
- Government and public records — statutory archives
- Healthcare — long-term patient records
- Finance and tax — documents that must be retained for years
- Libraries and academia — theses, journals, digital collections
For everyday documents — a flyer, a draft, a form you'll act on this week and forget — PDF/A is overkill. The larger file size and the restrictions (no encryption, no multimedia) buy you nothing if the document doesn't need to outlive its moment.
The Trade-offs
PDF/A's guarantees come at a cost:
- Bigger files, because every font and color profile is embedded rather than referenced.
- No password protection, as noted — encryption is off the table.
- It isn't automatic. A normal PDF is not PDF/A; you have to explicitly export or convert to it, and ideally *validate* the result. Free validators like veraPDF (the standard's official checker) confirm whether a file genuinely conforms, rather than just claiming to.
Common Misconceptions
"PDF/A is a different format I need special software to open." No — it's a PDF. It opens anywhere. The rules only affect what goes into it.
"Any PDF is fine for archiving." No. An ordinary PDF can reference external fonts, use encryption, or embed video — all of which threaten whether it will render correctly years from now.
"PDF/A guarantees my content is correct or legally valid." No. It guarantees faithful, self-contained *reproduction*. It says nothing about whether the content itself is accurate or legally sufficient.
The Bottom Line
PDF/A is PDF with a single-minded purpose: to still be readable, and to look exactly the same, decades from now. It gets there by embedding everything (fonts, colors), banning anything fragile (encryption, JavaScript, external links), and requiring self-describing metadata. Use it when a document genuinely needs to last — legal, medical, financial, government, academic records — and skip it when you're just sharing something for the moment. And whenever archiving matters, remember to convert deliberately and validate the result, because a regular PDF only becomes a lasting one on purpose.