Why can’t it add tags automatically?

Tags describe a document’s real structure: which parts are headings, paragraphs, lists, tables and images, and the order to read them in. Tags guessed by software are often wrong and make screen readers read in the wrong order or skip content — worse than no tags. The reliable route is to use proper heading styles in Word, InDesign or similar and export with “tagged PDF / accessibility” enabled, or to review the tags in Acrobat Pro.

If everything passes, is my PDF PDF/UA compliant?

Not necessarily. The tool covers what can be checked automatically; whether alt text is accurate, the reading order makes sense, colour contrast is sufficient and table headers are right still needs a human. Use the report as a to-do list, then verify with a dedicated checker such as PAC and a real screen reader.

What about scanned documents?

Scanned pages are only images, so a screen reader gets no text at all. Add a text layer with Scanned PDF OCR first, then come back to set the language and title; full PDF/UA compliance also requires tagging the OCR result in professional software.

Do the fixes change my pages? Is my file uploaded?

No and no. Fixes only write metadata — document language, title, window title setting, field tooltips, link descriptions and page tab order — and leave every word and image on the page untouched. Checking and fixing happen entirely in your browser. Encrypted files must be unlocked with Unlock PDF before they can be fixed.

♿

♿PDF Accessibility Checker - PDF/UA Tags, Language, Alt Text

More and more governments, universities and companies require published PDFs to work with screen readers (PDF/UA, WCAG, Section 508). This tool checks the parts of PDF/UA that can be verified automatically: whether the file is tagged (has a structure tree and declares Marked) and whether the tags are trustworthy; whether a document language (/Lang) and title are set and the window shows the title; whether every font is embedded and maps back to Unicode, so copying and read-aloud don’t produce gibberish; whether every Figure in the structure tree has alternate text; whether H1–H6 headings exist in a consistent hierarchy; whether page content is linked to the structure tree; the tab order on pages with annotations; whether form fields have tooltips (/TU); whether links have descriptions and Link tags; whether there are image-only scanned pages; and whether an encrypted file still allows accessibility extraction. Each item is marked pass, warning or fail with an explanation. Problems that can be fixed without risk are fixed for you: set the document language (suggested from the text), set the title and show it in the window, generate field tooltips from field names (editable one by one), describe links using the text under them, set structure tab order on tagged documents, and honestly re-mark a file that claims to be tagged but has no structure tree. The tool never invents tags — structure that isn’t real only misleads screen readers. Download the fixed PDF together with an HTML before / after report; everything runs in your browser.

Free · No sign-up · Runs in your browserLast updated: 2026-10-26

Loading tool...

How to use PDF Accessibility Checker

1Upload a PDF and review every accessibility check with its explanation
2Under Safe fixes, confirm the document language, title, field tooltips and link descriptions
3Click Apply fixes & download to get the fixed PDF and the report

FAQ

Why can’t it add tags automatically?: Tags describe a document’s real structure: which parts are headings, paragraphs, lists, tables and images, and the order to read them in. Tags guessed by software are often wrong and make screen readers read in the wrong order or skip content — worse than no tags. The reliable route is to use proper heading styles in Word, InDesign or similar and export with “tagged PDF / accessibility” enabled, or to review the tags in Acrobat Pro.
If everything passes, is my PDF PDF/UA compliant?: Not necessarily. The tool covers what can be checked automatically; whether alt text is accurate, the reading order makes sense, colour contrast is sufficient and table headers are right still needs a human. Use the report as a to-do list, then verify with a dedicated checker such as PAC and a real screen reader.
What about scanned documents?: Scanned pages are only images, so a screen reader gets no text at all. Add a text layer with Scanned PDF OCR first, then come back to set the language and title; full PDF/UA compliance also requires tagging the OCR result in professional software.
Do the fixes change my pages? Is my file uploaded?: No and no. Fixes only write metadata — document language, title, window title setting, field tooltips, link descriptions and page tab order — and leave every word and image on the page untouched. Checking and fixing happen entirely in your browser. Encrypted files must be unlocked with Unlock PDF before they can be fixed.