How to Check Accessibility in PDFs: Tools, Tests and Common Issues

This guide explains how to check accessibility in PDFs using automated tools and manual tests. It covers standards, a 10-minute audit, screen reader testing, common issues and a pre-publication checklist.

How to Check Accessibility in PDFs: a quick summary

How to Check Accessibility in PDFs starts with a mix of automated tests and a few manual checks. Run a machine check first, then confirm structure, reading order, alternative text and how the file behaves with a screen reader.

Which standards and rules to use

Target PDF/UA (ISO 14289) and WCAG 2.1 level AA as the primary technical standards. For public organisations check applicable local rules such as the UK Public Sector Bodies (Websites and Mobile Applications) Accessibility Regulations and the Equality Act for general duties on access. Also consider Section 508 if you work with US federal audiences.

Tools you should have

Adobe Acrobat Pro (Accessibility tool and Tags panel)
PAC 3 (PDF Accessibility Checker)
CommonLook PDF Validator or other enterprise validators
NVDA (free Windows screen reader), VoiceOver (macOS), JAWS (commercial)
OCR tool for scanned PDFs (Acrobat has built-in OCR)

Quick 10-minute audit: step-by-step

Open the PDF and confirm it is not image-only. If you cannot select text, run OCR.
Run an automated checker (PAC 3 or Acrobat Full Check). Note failures to fix later.
Check the Tags panel to confirm the document has a tag tree and heading tags where appropriate.
Confirm reading order visually and with the Reading Order tool or Tags order. Make sure content reads left-to-right, top-to-bottom.
Verify image alt texts in the Tags panel or properties for each figure and decorative marking.
Test a screen reader with a short pass: title, headings list, a paragraph and an image alt text.
Check forms and tab order so each form control has a programmatic label and logical tab sequence.
Confirm document language and metadata and that there are bookmarks for long documents.

Automated checks: what they catch and what they miss

Automated tools find missing tags, empty alt attributes, unlabelled form fields and PDF/UA rule failures. They do not reliably check reading meaning, whether alt text is descriptive, or whether table markup properly represents header relationships. Treat automated results as a prioritized bug list, not proof of accessibility.

Run an automated check in Adobe Acrobat Pro

Open Tools and choose Accessibility.
Choose Full Check (Accessibility Checker) and run against PDF/UA or custom profile.
Review the Accessibility Report and jump to flagged items in the Tags and Content panels.

Run PAC 3

Open the file in PAC 3 and run the validation. PAC 3 reports PDF/UA failures and gives a clear pass/fail list you can export for developers.

Manual checks you must do

Document structure: Confirm H1, H2 and other headings are marked as headings in the tag tree.
Reading order: Use the Reading Order tool or Tags order and simulate a screen reader read-through.
Alt text: Every informative image must have descriptive alt text. Decorative images should be marked as artifact or have empty alt text.
Tables: Data tables must have header cells and a correct scope. Avoid using tables for layout.
Forms: Every input needs a programmatic label. Check tooltips do not substitute for labels.
Document language: Set the primary language so screen readers use correct pronunciation rules.
Contrast and colour: Check contrast for text inside images and for graphical text; automated contrast tools may miss text inside pictures.
Security settings: Ensure the PDF allows screen readers to copy or access content. Remove restrictive permissions before publishing.

Use NVDA on Windows or VoiceOver on macOS for a quick validation. Steps to follow:

Start the screen reader and open the PDF.
Listen to the document title and first heading to confirm document language and structure.
Navigate by headings to confirm the heading order matches visual order.
Move to images and listen to alt text or the announcement that an image is decorative.
Tab through form controls to test labels and focus order.

Common mistakes and how to avoid them

Relying only on automated tools. Fix the items they find, then test with a screen reader and check meaning.
Leaving scanned pages as images. Run OCR and verify that the extracted text is tagged correctly.
Using visual headings only. Convert visual styles to real heading tags in the tag tree.
Adding alt text in the visual editor only. Ensure alt text appears in the tag tree or image properties used by assistive tech.

How to prioritise fixes

Fix issues that block access first: missing text (scanned images), unreadable navigation (no headings), unlabeled form controls, and protective permissions. Next handle informative issues such as long or vague alt text. Finally apply improvements like bookmarks and enhanced metadata.

Final pre-publication checklist

PDF text is selectable or OCR is applied
Tags exist and headings are correct
Reading order matches visual order
All informative images have alt text
Tables use header cells and logical structure
Forms have programmatic labels and correct tab order
Document language is set and metadata is present
File permissions do not block assistive tech
One screen reader pass completed