What is PDF/UA?
PDF/UA is an ISO standard (ISO 14289-1) that specifies what requirements a PDF document must meet to be universally accessible — hence the abbreviation UA (Universal Accessibility). The standard was first published in 2012 and builds on the general PDF standard (ISO 32000).
PDF/UA places requirements on the document’s internal structure: text must be machine-readable, headings must be tagged with correct structure tags (H1, H2, etc.), images must have alt texts, and tables must have correct header cells. A PDF/UA-compatible document must be able to be read and navigated meaningfully by a screen reader.
PDF/UA vs. WCAG 2.1
PDF/UA and WCAG 2.1 are two distinct standards with partially overlapping purposes. WCAG 2.1 is designed for web content and places requirements on HTML, CSS, and JavaScript-based solutions. PDF/UA is designed specifically for PDF documents and places requirements on the document’s internal tag structure, metadata, and embedded resources.
In practice, PDF/UA is the closest equivalent to WCAG for the PDF format. Guidance indicates that PDF files on public websites should meet PDF/UA — or alternatively be accompanied by an accessible HTML version. Both approaches are acceptable.
Core Requirements of PDF/UA-1
PDF/UA places numerous technical requirements on the document’s structure. The most important include:
- All pages must contain a tag tree with correct structure tags (P, H1-H6, Table, TR, TD, TH, Figure, Caption, etc.).
- All text must be included in the tag tree — no text may appear outside the structure.
- Figures and images must have alt text or be marked as artefacts if decorative.
- Tables must have correct header cells with scope attributes.
- The document must have a title in its metadata (XMP and DocInfo).
- The natural language must be specified in the document’s metadata.
- All links must have a descriptive alternative text.
Screen Readers and PDF/UA Support
Not all PDF readers support PDF/UA’s structural reading. Adobe Acrobat Reader has the best support for PDF/UA and accessibility tags — but even here the user experience for screen reader users is often far worse than browsing well-structured HTML.
Screen readers such as NVDA, JAWS, and VoiceOver can read PDF/UA-compatible documents, but navigation in PDFs is generally more complex and less predictable than navigating HTML pages. This is one of the reasons why many accessibility experts prefer HTML alternatives over making PDFs accessible.
How to Create PDF/UA-Compatible Documents
The easiest way to create PDF/UA-compatible documents is to start in a programme that supports export to tagged PDF — such as Adobe InDesign, Microsoft Word (with correct use of heading styles), or LibreOffice Writer. It is crucial that the document is created with correct style hierarchies from the outset.
A document correctly created in Word with heading styles (Heading 1, Heading 2, etc.) can be exported to a tagged PDF by selecting “Create PDF/XPS” and checking “Document structure tags for accessibility”. The result may not meet all PDF/UA requirements, but provides a far better starting point than an untagged PDF.
From PDF to HTML: When is Conversion the Best Choice?
In many cases, it is more efficient and user-friendly to convert an existing PDF to HTML than to attempt to fix its accessibility problems. The HTML format offers far better responsiveness, user-controlled scaling, and screen reader compatibility than even the best PDF/UA document.
PDFAccess is built for exactly this scenario. Whether your PDF is born-digital or scanned, PDFAccess converts it to semantic, WCAG 2.1 AA-compatible HTML — directly in your browser, for free, and without sending data to a server. For public bodies and libraries with large PDF archives, this is a practical and quick first step towards digital accessibility.