PDF vs PDF/A: The Complete Archival Guide
Written and fact-checked against our production PDF engine (pdf-lib, pdfjs-dist, and in-browser OCR where relevant). We ship the same code users run in the 24 EssentialPDFTool utilities—no paywalled “writer” edition.
You’ve been asked to deliver documents in “PDF/A format” and now you’re wondering what that means, why it matters, and whether your existing PDFs qualify. This guide cuts through the ISO standard jargon and gives you a practical understanding of PDF/A — when it’s required, what it actually prevents, and how to create or convert to it without expensive desktop software.
What Is PDF/A?
PDF/A is a subset of the PDF specification standardized by ISO (ISO 19005) specifically for long-term archival. The “A” stands for Archive.
The core problem PDF/A solves: a standard PDF can depend on external resources to render correctly. It can reference fonts that live on a server, embed JavaScript that requires a runtime, use encryption that could become inaccessible, or contain color profiles that reference external ICC files. Open that PDF in 30 years and you may find it renders differently — or not at all — because those external dependencies no longer exist.
PDF/A solves this by requiring complete self-containment. Every resource needed to render the document identically must be embedded within the file itself. Nothing may be fetched from outside. Nothing may execute code. The document must be a closed, deterministic artifact.
PDF/A Conformance Levels
The standard has evolved through three generations, each adding requirements:
PDF/A-1 (ISO 19005-1, 2005)
Based on PDF 1.4. The most restrictive version:
- All fonts must be embedded (including system fonts)
- No encryption of any kind
- No JavaScript, no embedded media, no transparent objects
- Color spaces must be fully specified (no device-dependent colors without profiles)
- XMP metadata required
Sub-levels:
- PDF/A-1a (Level A): Full accessibility conformance. Document must have logical reading order, tagged structure (headings, lists, tables), and Unicode mappings for all text. Screen readers can navigate it.
- PDF/A-1b (Level B): Basic visual reproduction. The document must look the same when rendered, but structural tagging is not required. Simpler to achieve.
PDF/A-2 (ISO 19005-2, 2011)
Based on PDF 1.7 (ISO 32000-1). Relaxes some restrictions and adds features:
- Allows JPEG 2000 image compression (more efficient than JPEG for lossless)
- Permits transparency (properly defined)
- Allows embedding of other PDF/A files (useful for attachments)
- Optional content (layers) is permitted
- Better support for digital signatures
Sub-levels: -2a (accessible + conforming), -2b (basic visual), -2u (Unicode — text extraction guaranteed but structure not required).
PDF/A-3 (ISO 19005-3, 2012)
Identical to PDF/A-2 but allows embedding of arbitrary file formats as attachments. You can embed an XML source file, a CSV, or even an original Word document inside a PDF/A-3. The embedded file doesn’t need to be PDF/A-conformant itself.
This makes PDF/A-3 popular for e-invoicing (ZUGFeRD in Germany, Factur-X in France) where human-readable and machine-readable versions coexist in the same file.
PDF/A-4 (ISO 19005-4, 2020)
Based on PDF 2.0 (ISO 32000-2). Simplifies the conformance model — removes the a/b/u sub-levels in favor of a single conformance with optional enhancements. Adds support for digital signatures (PAdES), better encryption handling, and associated files.
What Standard PDF Can Do That PDF/A Cannot
When converting to PDF/A, these features will be stripped or flagged:
| Feature | Standard PDF | PDF/A |
|---|---|---|
| JavaScript | Allowed | Prohibited |
| Encryption / passwords | Allowed | Prohibited (PDF/A-1/2/3) |
| External font references | Allowed | Prohibited (must embed) |
| Linked images/files | Allowed | Prohibited |
| Audio/video media | Allowed | Prohibited (PDF/A-1/2) |
| Transparency | Allowed | Prohibited (PDF/A-1) |
| Device-dependent colors | Allowed | Prohibited |
If your PDF contains a password, you cannot make it PDF/A-compliant without removing the password first. If it embeds a video, that must be removed. This is why archival conversions sometimes require decisions about what to preserve.
When You Must Use PDF/A
Legal and regulatory submissions
Courts in many jurisdictions require PDF/A for electronic filing. The US federal courts (PACER/CM-ECF), EU courts, and most national patent offices specify PDF/A-1a or PDF/A-1b. Using a standard PDF risks rejection.
Government and public records
Archival standards for government documents — including NARA in the US, the UK National Archives, and EU INSPIRE — specify PDF/A for permanent records. Local authorities often inherit these requirements.
Healthcare records
Many Electronic Health Record (EHR) systems export to PDF/A. Regulations like HIPAA in the US don’t explicitly require PDF/A but do require long-term retrievability, which PDF/A achieves.
Financial reporting
XBRL filings, annual reports, and fund prospectuses submitted to securities regulators (SEC, ESMA, FCA) increasingly require PDF/A.
Library and institutional archives
Universities, libraries, and cultural institutions use PDF/A for digital preservation programs. The Library of Congress recommends PDF/A-1 as a preferred format for textual documents.
E-invoicing
ZUGFeRD 2.x and Factur-X standards for electronic invoices use PDF/A-3 as the base format. This is becoming mandatory in Germany, France, and increasingly across the EU.
When Standard PDF Is Better
PDF/A isn’t always the right choice:
- Interactive forms: PDF/A-1 prohibits JavaScript, which many complex forms require. Use standard PDF.
- Encrypted documents: If confidentiality is required, standard PDF with AES-256 encryption is necessary.
- Rich media: Training videos, product demos embedded in PDFs need standard format.
- Smaller file size: Embedding all fonts and color profiles increases file size. For quick sharing, standard PDF compresses better.
- Web-optimized PDFs: Linearized (fast-loading) PDFs for web delivery are typically not PDF/A.
How to Identify If a PDF Is PDF/A
Using EssentialPDFTool
Open your PDF in any of our tools. The metadata panel (when available) will show the pdfaid:conformance and pdfaid:part XMP metadata fields that indicate PDF/A compliance.
Manual check
PDF/A-compliant files contain XMP metadata with the following namespace:
xmlns:pdfaid="http://www.aiim.org/pdfa/ns/id/"
And properties like:
<pdfaid:part>1</pdfaid:part>
<pdfaid:conformance>B</pdfaid:conformance>
Open the PDF in a text editor (it’s partly plain text) and search for pdfaid to find these markers.
Validator tools
Free PDF/A validators include veraPDF (the reference implementation supported by the PDF Association) and Adobe Acrobat’s Preflight panel.
Converting to PDF/A
Our PDF to PDF/A converter handles conversions entirely in your browser:
- Upload your PDF. The tool inspects it for common incompatibilities.
- Select conformance level. For most uses, PDF/A-2b offers the best balance of compatibility and features.
- Convert. The tool embeds fonts, converts colors to sRGB where needed, removes JavaScript, strips encryption, and writes the required XMP metadata.
- Download. Your PDF/A file is ready for submission.
For complex PDFs with transparency or multimedia, review the output in a validator like veraPDF before submitting to a regulatory body.
Common Conversion Pitfalls
Font licensing. Some fonts prohibit embedding. If a font can’t be embedded, conversion fails or substitutes a generic font. Always use open-license fonts (like the Google Fonts library) in documents destined for archival.
Device-dependent colors. RGB colors defined without an ICC profile are device-dependent — the same values look different on different screens. PDF/A requires all colors to be tagged with a profile (typically sRGB IEC 61966-2-1 for RGB content or ISO Coated v2 for CMYK print). A converter should handle this automatically.
Transparency flattening. PDF/A-1 prohibits transparency. Converting a PDF with transparent logos or drop shadows requires flattening those layers into opaque composites. This is lossy for editing purposes but visually transparent (pun intended) in the output.
Metadata requirements. PDF/A requires a properly formed XMP metadata block. If your PDF’s metadata is malformed or missing required fields, the converter must generate or repair them.
File Size Impact
PDF/A files are typically 10–30% larger than equivalent standard PDFs because:
- All fonts must be embedded (not referenced)
- Color profiles are embedded
- Metadata is more extensive
- No compression tricks that might compromise reproducibility
For long-term archival, this is an acceptable tradeoff. For frequent sharing or web delivery, keep a standard PDF and convert to PDF/A only for regulatory submissions.
Quick Decision Guide
| Situation | Use |
|---|---|
| Court filing, patent submission | PDF/A-1b minimum, PDF/A-1a if accessibility required |
| Government archive | PDF/A-2b (modern, good compression) |
| E-invoice | PDF/A-3b (allows XML attachment) |
| Password-protected document | Standard PDF |
| Interactive form | Standard PDF |
| General sharing, email | Standard PDF |
| Long-term institutional archive | PDF/A-2a or PDF/A-3a |
PDF/A is not complexity for its own sake — it’s a guarantee. A document that conforms to the standard will render the same way in 2055 as it does today, on whatever software exists then. For records that matter, that guarantee is worth the tradeoff.
Share this article
Try these free tools