Skip to main content

How to redact a PDF using Microsoft Word

Redact in Word, export to PDF — and run the Document Inspector so comments, edit history, and metadata don't leak.

  • Microsoft Word

  • Mac & Windows

  • 6 min read

  • Last verified

  • By Edward Coleridge Smith

On this page
All guides

TL;DR

  • You need the original .docx, not the PDF — Word can't redact a PDF directly
  • Replace sensitive text with [REDACTED] (or similar). Don't highlight in black — the text underneath survives the PDF export
  • Accept all tracked changes and turn track changes off before exporting
  • Run Inspect Document — it strips comments, author info, and revision history
  • Export with File → Save As → PDF (or Export → PDF)
  • Verify with our /verify page before sharing

Why most people get this wrong

Word is a perfectly good tool for producing a redacted PDF — provided you have the original .docx. Two failure modes. First: highlighting text in black and exporting — the underlying characters survive and can be selected straight out of the PDF. Second: forgetting to clear track changes, comments, and document metadata — the exported PDF then leaks author names, edit history, and reviewer commentary.

Both are easy to avoid once you know to look for them.

Open the source document in Word

Open the original .docx (or .doc) in Microsoft Word. If you only have a PDF, Word can't redact it directly — convert it to a Word document first (File → Open with a PDF in newer versions does this automatically) or use Acrobat instead.

Word with a document open

Replace sensitive text with redaction marks

The safest approach is destructive replacement: select the sensitive text and replace it with [REDACTED], █████, or similar. For repeated names or numbers, use Edit → Find → Replace (or Cmd+H / Ctrl+H) to do it in one pass. Don't apply a black highlight or font colour — both keep the underlying text intact.

Word's Find and Replace dialog

Accept all changes and turn off track changes

If track changes is on, the redacted versions will appear as edits and the original will be preserved as the "before" state. Both will be exported into the PDF. Go to Review → Track Changes and turn it off, then Review → Accept → Accept All Changes.

Word's track changes panel showing all changes accepted

Run the Document Inspector

Word stores comments, author information, hidden text, custom XML, and revision history alongside the visible document. File → Info → Check for Issues → Inspect Document (Windows) or Review → Protect → Remove Personal Information (Mac) opens the inspector. Run it and remove everything it flags except what you want to keep.

Word's Document Inspector dialog with issues found

This is the step that prevents Reviewed by John Smith from ending up in your redacted PDF's metadata.

Export to PDF and verify

File → Save As (or File → Export), choose PDF as the format, and save under a new filename. Don't tick "Save as PDF/A" unless your archival workflow requires it — PDF/A imposes constraints you probably don't need.

Word's Save As dialog with PDF selected

Open the exported PDF in a fresh window. Try to select any text where redactions were applied (you should only get the replacement characters). Search for one of the redacted strings (no results). For an automated check, drop it into /verify.

When to automate this

Doing this for one document is fine.

Doing it for thousands, or letting your users do it inside your product, is what the Redactr API is for.

See plans

FAQ

Can I redact directly inside a PDF using Word?

Not effectively. Word can open a PDF and convert it to an editable document, but layout often drifts. For PDF-only inputs, Acrobat is the better tool.

What about Word for the web?

Word on the web (Office 365 in a browser) doesn't expose the Document Inspector. Use the desktop app for the final pass.

Is highlighting text in black good enough?

No. A black highlight is a visual property — the text underneath stays in the file and is selectable in the exported PDF.

Do I need to worry about embedded fonts or comments after exporting?

Comments are stripped by the Document Inspector. Embedded fonts aren't usually a leak risk in themselves, but the inspector handles author and revision metadata that often matters more.