PDF Accessibility Best Practices (WCAG, Tagged PDFs, Alt Text)

TL;DR

Accessible PDFs are tagged PDFs — every element has a logical role (heading, paragraph, table, figure with alt text). Author accessibly in Word, then export. For existing PDFs, use Acrobat Pro's 'Make Accessible' wizard and validate with PAC 2024. Tagged PDFs help screen readers, search engines, and AI parsers alike.

Why accessibility is no longer optional

Accessibility laws (ADA in the US, EAA in the EU, AODA in Canada) increasingly apply to PDFs published by businesses, governments, and educators. Beyond compliance, accessible PDFs are also better understood by Google, AI search engines, and document AI pipelines — the same tags that help screen readers help every machine reader.

What 'accessible' actually means

An accessible PDF is a 'tagged PDF': behind the visual layout, every element has a logical tag — H1, H2, paragraph, list, table, figure with alt text. Tags give screen readers a logical reading order, let users navigate by heading, and make tables make sense to non-visual users. PDF/UA (ISO 14289) and WCAG 2.2 are the standards to aim for.

Start in the source, not the PDF

The cheapest accessible PDF is one that was authored accessibly. In Word: use built-in heading styles (don't fake headings with bold + bigger font), add real lists (not '* ' typed manually), add alt text to images, set table headers properly. When you export to PDF (File → Export → 'Best for electronic distribution and accessibility'), Word carries all the structure into the PDF as tags. Convert your accessible Word doc with our Word to PDF tool.

What to fix in the PDF itself

If you can't go back to the source, Acrobat Pro's 'Make Accessible' wizard handles 80% of the work: it runs OCR, adds tags, prompts for a document title and language, and asks for alt text on every image. The remaining 20% is manual: confirming reading order, fixing wrongly-detected headings, and writing meaningful alt text (not 'image001.jpg').

Tables and forms need extra care

Tables: mark header rows so screen readers announce 'Column: Total' before each cell. Avoid merged cells when possible. Forms: every field needs a tooltip (which becomes the screen-reader label), tab order must follow visual order, and required fields must be marked.

Test with real tools

Don't ship without testing. Acrobat Pro's accessibility checker catches structural issues; the free PAC 2024 (PDF Accessibility Checker) goes deeper. For real-world testing, use a screen reader yourself — NVDA on Windows is free, VoiceOver ships with macOS. If you can navigate the PDF with your eyes closed, your users probably can too.

Common myths

'Saving as PDF/A makes it accessible' — false. PDF/A is about archival, not accessibility. PDF/A-2a and PDF/UA together are the right combo for accessible archives. 'OCR makes a scanned PDF accessible' — partially. OCR adds text, but you still need tags, reading order, and alt text. 'Accessibility hurts SEO' — the opposite. Tagged PDFs rank better in Google because the tag tree gives semantic structure to indexers.

Frequently asked questions

What's the difference between PDF/A and PDF/UA?

PDF/A is for long-term archiving (self-contained, no JavaScript). PDF/UA is for accessibility (tagged structure, reading order, alt text). They're complementary — PDF/A-2a + PDF/UA together is the gold standard.

Can I make an accessible PDF for free?

Yes — author with proper styles in Word or LibreOffice, export to PDF. For existing PDFs, the free PAC 2024 checker tells you what's missing. Adobe Acrobat Pro is the easiest way to fix tags but is paid.

Does Google care about PDF accessibility?

Yes. Tagged PDFs give Google semantic structure (headings, lists, tables) which improves indexing and ranking. Untagged image-only PDFs are nearly invisible.

How do I add alt text to a PDF?

In Acrobat Pro: open the Tags panel, find the <Figure> tag, right-click → Properties → Alternate Text. In Word before export: right-click any image → Edit Alt Text.

Are scanned PDFs accessible?

Not by default — they're image-only. Run OCR first (learn how), then add tags and reading order. Acrobat's 'Make Accessible' wizard does both in one workflow.