PDF Is Not Readable by Screen Readers — How to Fix Accessibility
PDFs that are silent or jumbled when read by a screen reader are untagged or incorrectly tagged. Learn how to make any PDF accessible to assistive technology.
A PDF that a screen reader like NVDA, JAWS, or VoiceOver cannot read — either producing silence, random character sequences, or a completely wrong reading order — is either an image-based PDF with no text layer, an untagged PDF with no structure information, or a tagged PDF with structural errors. Each problem has a different fix, and the right one depends on which type you have.
Step 1: Identify What Type of PDF You Have
Three scenarios: (1) Image-based PDF (scanned document) — no text layer at all. Screen reader reads nothing. Fix: add OCR text layer. (2) Untagged PDF (has text but no structure tree) — screen reader reads text in display order, which is often wrong (columns, headers, footers mixed in). Fix: add accessibility tags. (3) Tagged PDF with errors — screen reader reads something but in wrong order, skips content, or reads alternative text incorrectly. Fix: remediate the tag tree. To check: in Acrobat Reader, File → Properties → Description → look for "Tagged PDF: Yes" or "No."
Fix for Image-Based PDFs: Add OCR
A scanned PDF has no text for a screen reader to read. The fix: run OCR to add a searchable text layer. Use FixMyPDF OCR to add a text layer in your browser — select the correct language for the document, which significantly affects accuracy. After OCR, a screen reader can read the recognized text. The quality depends on scan resolution and font clarity: a clean 300 DPI scan of printed text achieves 98-99% accuracy; faded or handwritten documents may need manual correction for critical accessibility.
Fix for Untagged PDFs: Auto-Tag in Acrobat
In Acrobat Pro: Tools → Accessibility → Add Tags to Document. Acrobat analyses the layout and generates a structure tree automatically. For simple documents (single column, standard heading hierarchy), auto-tagging produces good results. For complex layouts (multi-column, tables, sidebars), auto-tagging creates a tag tree that needs manual review and correction. After auto-tagging, run the Accessibility Checker (Tools → Accessibility → Accessibility Check) to identify remaining issues.
Fix Reading Order Issues
Even in tagged PDFs, the reading order in the structure tree may not match the visual reading order — especially for multi-column layouts where columns were tagged left-to-right across both columns rather than down each column independently. In Acrobat Pro: Tools → Accessibility → Reading Order tool. This shows the numbered reading order of all content blocks. Drag blocks to reorder them, or use the Tags panel to restructure the tree. Fix the order so it reads: document title → navigation → column 1 top-to-bottom → column 2 top-to-bottom.
Critical: Alternative Text for Images
Every image, chart, and diagram must have meaningful alternative text for screen readers. In Acrobat Pro: right-click an image → Edit Alt Text. Write a description of what the image conveys (not just "chart" — say "bar chart showing quarterly sales increasing from $2M in Q1 to $3.5M in Q4 2025"). Decorative images (dividers, background textures) should be marked as Artifacts rather than given empty alt text. In the Tags panel, right-click decorative elements → Change Tag to Artifact.
Validate With a Screen Reader
After tagging, test with the actual tools: download NVDA (free, Windows) or use VoiceOver (built into macOS and iOS). Open the PDF and listen. The reading should: start with the document title, follow headings logically, read table contents row by row with column headers, and skip decorative elements. Fix any content that reads incorrectly. For government and regulated-industry PDFs, also run the PAC accessibility checker (free from access-for-all.ch) which provides the most thorough PDF/UA validation available.
Try PDF to Text Now — Free
Browser-based, private, and instant. No account or software required.
Open PDF to Text