How to remediate PDFs

PDF remediation improves accessibility and usability for people who use assistive technology. This guide summarises practical steps and best practices to convert inaccessible PDF content into a structured, navigable and testable document.

Plan before you start

Good remediation begins with preparation. Confirm the target accessibility standard (for example national or WCAG requirements), gather source files where possible, and identify priority documents and deadlines. Whenever you can, work from the original source (Word, InDesign, etc.) rather than remediating a flat, scanned PDF.

Key remediation steps

1. Set document metadata and language

• Give the document a meaningful title and ensure metadata fields are complete.
• Declare the primary language so screen readers use the correct pronunciation.

2. Establish a logical structure with tags

• Create or repair the tag tree so content follows a semantic order: headings, paragraphs, lists, tables and figures should be tagged appropriately.
• Use correct heading levels in sequence (no skipping from heading level 1 to 3) and verify list items are nested as lists, not plain paragraphs with bullets.

3. Ensure correct reading order and reading pane

• Check the artefacts and content order; reading order should match visual order for linear navigation by a screen reader.
• Remove hidden or decorative content from the tag tree and mark purely decorative elements as artefacts where possible.

4. Images, figures and alternative text

• Provide concise, descriptive alt text for informative images. For complex images, include a longer description or refer to a text alternative in the document body.
• Mark purely decorative images so they are ignored by assistive technology.

5. Tables and data

• Tag tables as tables and mark header cells appropriately. Avoid using tables for layout.
• Provide summaries or captions when tables are complex and ensure column and row headers are identified.

6. Forms and interactive content

• Use accessible form fields with programmatic labels and correct field types. Ensure tab order is logical and that instructions are available to assistive technology.
• Verify that form controls are operable with keyboard and that required fields and errors are exposed to assistive technology.

7. Scanned documents and OCR

• Run optical character recognition (OCR) on scanned pages to produce selectable, searchable text before tagging.
• Review OCR output carefully for accuracy and correct reading order; misrecognised text must be corrected manually.

Testing and validation

• Use automated accessibility checkers to catch common issues, then perform manual checks to address things automation misses.
• Test with one or more screen readers and keyboard-only navigation to confirm the document is usable in real scenarios.
• Validate against the chosen accessibility standard and iterate until errors and warnings are resolved or explained.

Practical remediation workflow

1. Gather original source files and assets where possible.
2. Run OCR on scanned documents and correct text output.
3. Create or repair the tag structure, headings and lists.
4. Add alt text, form labels and table headers.
5. Set language and document metadata.
6. Perform automated checks, then manual testing with a screen reader and keyboard navigation.
7. Fix issues discovered during testing and re-check until acceptable.

Quick checklist

• Document title is meaningful and language is set.
• All structural elements are tagged correctly (headings, lists, paragraphs).
• Images have appropriate alt text or are marked decorative.
• Tables are tagged and header cells identified.
• Form fields have labels and logical tab order.
• Scanned pages have accurate OCR text.
• No security settings block assistive technologies.
• Manual testing with assistive technology has been completed.

Common pitfalls and fixes

• Unordered or missing tags: rebuild the tag tree and set appropriate roles.
• Visual headings that are not tagged: convert visual styles into semantic heading tags.
• Incorrect table markup: re-tag cells and designate header relationships.
• Forms without labels: attach programmatic labels and add descriptive text for complex fields.
• Poor OCR accuracy: correct recognised text and re-run OCR settings or use the original source document.

Maintainability and training

Document accessible authoring guidelines and train content creators to produce accessible source files. Where possible, embed accessibility checks into the publication workflow so remediation is faster and fewer issues reach the final PDF.

Final recommendations

Combine automated tools with manual inspection and real-world testing. Prioritise high-impact documents, keep a clear remediation checklist, and update processes so accessibility is part of routine content production rather than an afterthought.