PDF/A Explained: When You Need the Archival Format (And When You Don't)

Every few weeks someone asks us why their court clerk, government portal, or compliance officer rejected their PDF. Nine times out of ten the answer is the same: the system wanted PDF/A, and what they uploaded was plain PDF. The file looked identical. The rejection email was terse. The deadline was tomorrow.

PDF/A is one of those specifications that the people who work with it every day treat as obvious and the rest of us have never heard of. It is not an alternative to PDF — it is a subset of it, a stricter version designed for long-term preservation. This post explains what PDF/A actually is, why it exists, when you genuinely need it, and when plain PDF is perfectly fine.

What PDF/A is

PDF/A is the archival standard for PDF files, formalized as ISO 19005. The "A" stands for "archival." It was first published in 2005 and has been revised several times since, most recently as PDF/A-4 in 2020. The goal is simple: produce a PDF that will render faithfully in 50 or 100 years, on software that does not yet exist.

To guarantee that, PDF/A imposes restrictions on regular PDF. The big ones:

• All fonts must be embedded. A regular PDF can reference system fonts; a PDF/A cannot. If the font is not inside the file, the file is not compliant.
• No external dependencies. No hyperlinks to external files, no JavaScript, no multimedia that relies on external codecs, no attached files that need separate software to open.
• No encryption. Archives need to be openable decades from now without lost passwords.
• Standardized color. Colors must be specified unambiguously (device-independent), often via an embedded ICC profile, so a 2050 printer produces the same output as a 2026 one.
• Metadata in a specific structured format. XMP metadata, with known fields, so a future cataloging system can read it without reverse-engineering the file.

The underlying question PDF/A answers is: if a reader in 2090 tries to open this file with software no one has written yet, do they get the same document? Regular PDF does not promise that. PDF/A does.

Why the standard exists

Digital preservation is harder than it looks. A document format that works perfectly in 2026 can quietly degrade over decades as the surrounding software ecosystem changes. Fonts get removed from operating systems. Proprietary extensions get deprecated. Codecs stop shipping. Links rot. A modern PDF that references Helvetica and a website will, in 40 years, render with a font substitute and a broken link — subtly different from the original.

That is unacceptable for certain classes of document: legal filings that may be cited in future cases, medical records that must be retained by law, government archives that survive political transitions, historical scholarship that needs to be cited verbatim. For those uses, "probably looks right in the future" is not a good enough guarantee. PDF/A is the answer — a format that trades some flexibility for stronger long-term fidelity guarantees.

The PDF/A variants

Four versions exist, each slightly less strict than the last.

• PDF/A-1 (2005) is the strictest. Based on PDF 1.4. No transparency, no JPEG 2000, no layers. Sublevels: -1b (basic) and -1a (accessible).
• PDF/A-2 (2011) adds JPEG 2000, transparency, and layers. Sublevels: -2b, -2a, and -2u (Unicode-mapped text).
• PDF/A-3 (2012) allows arbitrary file attachments inside the container — used by the German and French e-invoicing standards (ZUGFeRD, Factur-X).
• PDF/A-4 (2020) collapses the earlier variants; adoption is still growing.

If you were told to submit "a PDF/A" with no variant specified, PDF/A-2b is a safe default — widely accepted, permissive enough for modern content, strict enough for most archival needs.

When you actually need PDF/A

The format is mandatory or strongly recommended in several real-world contexts:

• Legal filings. Many court e-filing systems require PDF/A (often -1a or -2a). Most European jurisdictions already require it.
• Government archives. National archives in the US, UK, and Germany require PDF/A for long-term preservation. The Library of Congress recommends it for federal records.
• Regulated industries. FDA submissions, FINRA-regulated correspondence, and HIPAA-compliant medical records retention commonly specify PDF/A.
• Academic repositories. Many university libraries require PDF/A for theses and dissertations.
• E-invoicing. ZUGFeRD and Factur-X are built on PDF/A-3.
• Long-term business archives. Contracts with 10+ year retention obligations often name ISO 19005.

If you have a clear external requirement for PDF/A — a portal that rejects non-compliant files, a compliance policy that names the standard — you need it.

When plain PDF is fine

Most documents, most of the time, do not need PDF/A. Regular PDF works perfectly for:

• Email attachments and day-to-day sharing. Your recipient will open this file in the next two weeks, not in 2075.
• Contracts you intend to re-sign or supersede. The working version will change; only the final signed copy matters, and that may or may not need archival format depending on your retention policy.
• Marketing materials, one-pagers, and web downloads. These have short useful lifespans by design.
• Internal reports and working documents. If they end up in a long-term archive, they can be converted later.

PDF/A also has practical tradeoffs that matter for non-archival use. Files are usually somewhat larger (all fonts embedded, no encryption, often no sophisticated compression). The format is less flexible (no JavaScript, no external links). And the conversion itself can subtly change a document — fonts that were system-referenced get re-embedded, and colors specified in screen-native RGB get converted to device-independent ICC equivalents. For a legal filing this is exactly what you want. For a brochure, it is overkill.

A useful rule of thumb: if someone is going to need to re-open this file more than five years from now, and nobody will be available to convert it then, use PDF/A now. Otherwise, regular PDF is fine and cheaper.

How to convert to PDF/A

Our PDF to PDF/A tool handles the conversion in your browser:

• Open PDF to PDF/A on PDF Genie.
• Drop your file in.
• Pick the target conformance level (PDF/A-2b is the safe default).
• Download the converted file.

Under the hood, the converter embeds any referenced fonts, strips disallowed features (JavaScript, encryption, external links), adds the required XMP metadata with the conformance declaration, and assigns device-independent color profiles. The output file opens as a regular PDF in any viewer but identifies itself as PDF/A-compliant when checked by tools that care.

Expect a modest file-size increase after conversion — typically 10-20% — driven almost entirely by font re-embedding (PDF/A requires every used font to be fully embedded). Validation is worth running: the open-source veraPDF validator will tell you whether your output is actually PDF/A-2b compliant, not just "PDF/A-shaped." Files with unusual features — JavaScript-driven form calculations, external hyperlinks, media attachments — have those features stripped during conversion. If your source file relies on them, the conversion removes them silently; open the output and verify the static content is intact before filing.

Validating the output

If someone is going to reject a non-compliant file at the other end, it is worth validating your output before submission. Two options:

veraPDF is the open-source industry-standard validator, maintained by the PDF Association. Free, reliable, slow.

Adobe Acrobat Pro's Preflight tool has a built-in PDF/A profile checker. Paid, fast, widely trusted by people who file regulatory documents for a living.

Our own tool runs a lightweight conformance check during conversion and flags obvious problems, but for high-stakes submissions we recommend running a second validator as well. Redundancy is cheap; a rejected filing is not.

Honest caveats

PDF/A conformance is a moving target. The original 2005 spec has been revised several times, readers interpret the rules inconsistently, and some validators disagree with each other on edge cases. Our converter aims at PDF/A-2b as the most widely accepted variant and has passed internal testing on several hundred real documents, but we make no claim that every conceivable source PDF will convert cleanly. If your submission portal rejects the output, send us the file — we triage these case-by-case and often learn from them.

Convert to PDF/A now

Convert your PDF to PDF/A — free →