Compliance teams face growing pressure to ensure that every digital document meets accessibility standards. PDFs remain one of the most common formats for contracts, reports, statements, and public filings, yet they are also among the hardest to make accessible. This checklist gives compliance professionals a clear, repeatable process for auditing PDFs against WCAG 2.1, Section 508, and PDF/UA requirements.
Why PDF accessibility matters for compliance
Accessible PDFs are not just a legal safeguard. They improve usability for employees, customers, and stakeholders who rely on screen readers, magnification, keyboard navigation, or speech recognition. Regulators in the United States, the European Union, Canada, and Australia now treat digital accessibility as a civil rights and consumer protection issue. A single inaccessible PDF can trigger complaints, litigation, or reputational damage. Building a checklist-driven audit process reduces that risk and demonstrates due diligence.
The ultimate PDF accessibility checklist
1. Document properties and metadata
Start with the document's foundation. Open the file's properties and confirm that a descriptive title is set, not just a filename. The title should match the document's purpose and appear correctly in assistive technology. Add an author, subject, and keywords that help users find and identify the file. Set the document language in the metadata so screen readers select the right voice profile.
2. Tags and structure
Tags provide the semantic backbone of an accessible PDF. Verify that headings use proper H1 through H6 hierarchy without skipping levels. Check that paragraphs, lists, and block quotes are tagged correctly. Remove empty tags and artifacts that do not convey meaning. Ensure that the document has a single top-level H1 and that sections follow a logical nested structure.
3. Reading order
Screen readers follow the tags tree, not the visual layout. Open the reading-order panel and confirm that content flows in the same sequence a sighted user would expect. Watch for multi-column layouts, sidebars, and captions that appear out of order. If a document uses tables or figures, make sure their associated text appears at the correct point in the reading sequence.
4. Alternative text
Every meaningful image, chart, diagram, and icon needs concise, accurate alternative text. Decorative elements should be marked as artifacts so screen readers skip them. For complex charts, provide a nearby text summary or long description. Avoid redundant phrases like "image of" or "picture of." Instead, describe the information the visual conveys.
5. Tables
Tables must use proper header cells, scope attributes, and consistent structure. Identify row and column headers so assistive technology can announce relationships correctly. Avoid merged cells whenever possible, because they break the predictable grid that screen readers rely on. Do not use tables purely for visual layout; use them only for genuine data relationships.
6. Forms and interactive fields
Interactive PDF forms need labeled fields, clear instructions, and a logical tab order. Each field should have a tooltip or accessible name that explains what information is required. Confirm that required fields are indicated with text labels, not color alone. Test that checkboxes, radio buttons, and dropdowns can be operated with a keyboard and that error messages are accessible.
7. Links and navigation
Links should use meaningful text rather than raw URLs. Verify that every hyperlink has an accurate destination and that bookmarks or a table of contents is provided for longer documents. Ensure that link text is unique within the page when it leads to different targets. Check that internal cross-references remain valid after the PDF is generated.
8. Language and fonts
Set the primary language for the document and mark any passages in a different language with the correct lang attribute. Use fonts that embed all needed characters and avoid relying on images of text. Confirm that text can be resized and reflowed without loss of meaning, and that there is sufficient contrast between text and background colors.
9. Testing and validation
Automated tools catch many issues, but manual testing is essential. Run the PDF through an accessibility checker, then open it with a screen reader such as NVDA or JAWS. Tab through the document using only a keyboard. Listen for headings, links, form labels, and table relationships. Document findings in a remediation log and re-test after fixes.
Closing thoughts
PDF accessibility is a team effort that spans content authors, designers, developers, and compliance officers. By following this checklist on every release, organizations can reduce legal exposure, improve user experience, and show a genuine commitment to inclusion. Start with the highest-risk documents, build the habit, and scale the process across your document library.
Related: Run the free PDF accessibility checker or read how to make a PDF ADA compliant.
Sources and references
- W3C — WCAG 2.1
- ISO 14289-1:2014 — PDF/UA-1
- veraPDF — PDF/UA Validation
- ADA.gov — Web Accessibility Guidance
Detailed checklist: what to verify, item by item
Structure and tag tree
- Does the document have a tag tree? (In Acrobat: View → Show/Hide → Navigation Panes → Tags. If the panel is empty, the document is untagged.)
- Are headings tagged as H1, H2, H3 in the correct nesting order? A document should not skip from H1 to H4.
- Are all content paragraphs tagged as P?
- Are lists tagged as L containers with LI children? Each bullet point should be an LI.
- Do decorative page elements (borders, background shapes, repeating headers/footers) appear in the tag tree? If so, they should be marked as artifacts — they are visual noise, not content.
Reading order
- Does the tag order match the visual reading order? Open the Tags panel and the Order panel side by side.
- In multi-column sections, does the order read down one column before moving to the next? If the validator shows cross-column reading, the order is broken.
- Are footnotes and endnotes placed after their referring content, not before?
Metadata
- Document title: File → Properties → Description → Title. It should not default to the filename.
- Language: File → Properties → Advanced → Language. Set to the correct BCP 47 tag (e.g., en-US, de-DE, fr-FR).
- Tagged PDF flag: File → Properties → Description → Tagged PDF: Yes. This must be explicitly true — it is the validator's first check.
Figures and alternative text
- Is every Figure tag associated with an Alt entry? If the image is decorative (background, ornament), mark it as an artifact instead.
- Does the alt text describe the information the figure conveys, not its appearance? "Q3 revenue, up 18% driven by new subscriptions" passes; "blue bar chart" does not.
- For complex figures (charts, diagrams, maps): does the alt text include a short description and a reference to a longer description or data table elsewhere in the document?
- See how to add alt text to a PDF for the full method.
Tables
- Does every data cell (TD) reference a header cell (TH) via the Headers or Scope attribute?
- Do tables have a caption or summary that describes what the table contains?
- Are layout tables (used for visual formatting, not data) marked as artifacts? Only real data tables belong in the tag tree.
Forms
- Does every form field have a Tooltip or linked label widget?
- Is the tab order logical (top to bottom, left to right)?
- Are required fields indicated in a way a screen reader can detect?
Links and navigation
- Do hyperlinks use meaningful display text rather than raw URLs or "click here"?
- Does the document have bookmarks for sections, matching the heading structure?