GUIDE · 19 NOVEMBER 2026

How to redact a PDF properly so hidden text is really removed

Learn how to redact a PDF so hidden text is truly deleted, which methods are unsafe, and how to test the saved file before you share it.

Last updated: 19 November 2026

To redact a PDF properly, use a tool that deletes the selected text and images from the file, then save a new copy and test it. Black boxes, highlights and white rectangles only cover content. The text underneath usually stays in the file and can be copied, searched or extracted.

This article covers which methods are safe, how to redact step by step, and how to check the result before you send it. For the wider picture of editing PDFs locally, see our guide to editing PDFs on a Mac without uploading them.

What is the difference between covering text and redacting it?

Covering text only changes how a page looks. Redaction changes what the file contains. Smallpdf describes redaction as the permanent removal of selected text or images so they can't be viewed or recovered, and says that blurring or covering text isn't true redaction because the text often remains under the effect.

A PDF is built in layers. The words on a page are stored as text objects. A black rectangle drawn with a "shape" or "markup" tool is a separate object placed above them. Anyone who opens the file in an editor, selects the area, or presses Cmd+A and copies into a text document sees the original words. PDFgear makes the same point: in a fake redaction the text is only hidden visually and can still be exposed.

Common fake redactions include:

  • A black or white rectangle drawn with a shape tool.
  • A yellow highlight changed to black.
  • Text changed to white on a white background.
  • A blur or pixelation effect applied in a viewer.
  • A black-and-white image of a page with a "hidden" layer still attached.

All of these leave the original data in the file. Real redaction deletes the data and then, usually, paints a box where it was.

Which redaction methods are safe, and which aren't?

A method is safe only if the original text no longer exists in the saved file. Look for a dedicated redact function that marks content and then applies the removal. Anything that only draws on top is unsafe.

MethodRemoves underlying text?Safe for sensitive files?
Black rectangle from a shape or markup toolNoNo
Highlight set to blackNoNo
White text or white boxNoNo
Delete text in a basic PDF editorOften, but not always (can leave copies in metadata, bookmarks or earlier saved versions)Only after testing
Dedicated redact tool that marks, then appliesYes, when appliedYes, after you verify
Print to PDF or screenshot, then OCRRemoves the original text layer, but the result is an image unless re-recognisedYes, if redaction is applied before flattening
Redact in the source document, then export a fresh PDFYesYes, usually the most reliable

The last row is worth remembering. If you still have the Word, Pages or spreadsheet file, deleting the sensitive text there and exporting a new PDF is often more reliable than editing the old PDF.

How do you redact a PDF step by step?

Work on a copy, mark everything you need to remove, apply the redactions, and save under a new name. Never overwrite your only original.

  • Make a working copy. Name it something like contract-REDACTED.pdf. Keep the original in a private folder.
  • List what must go. Names, addresses, account numbers, signatures, case numbers, email addresses, phone numbers, dates of birth. Write the list down first so you don't miss repeats.
  • Mark the content. Use the redact tool to select text or draw an area over images and scanned regions. If your tool supports search-based marking, search for each name or number so every occurrence is caught, including in headers, footers and footnotes.
  • Check each mark. Make sure the marks cover whole words and don't leave descenders or partial characters sticking out.
  • Apply the redactions. In most tools, marking is reversible and only applying removes the content. This is the step people forget, and it's why marked-but-not-applied files leak.
  • Remove metadata. Strip author names, titles, comments and document properties. PDFgear's guide mentions an optional "Sanitize Document" step in Acrobat Pro to remove metadata and hidden information before saving.
  • Save as a new file. Then run the tests below on that saved file, not the one open on screen.

What about scanned documents?

A scan is a picture, so there is no hidden text layer unless someone ran OCR on it. Redacting a scan means covering pixels permanently so the image itself is altered. If the PDF has been made searchable, there is also a text layer behind the image, and that layer must be redacted too. If you plan to run OCR on a scan, do it before redacting, so the new text layer is included in what you remove.

How can you check that a redaction really worked?

Open the saved file and try to get the text back the way a stranger would. If nothing comes out, you have good evidence the redaction worked. These tests take about two minutes.

  • Search test. Open the redacted file in Preview or another viewer, press Cmd+F and search for each redacted name and number. There should be zero matches.
  • Select-all test. Press Cmd+A, copy, and paste into a plain text editor. The redacted words should not appear.
  • Drag test. Click and drag across a black box. If text highlights beneath it, the redaction failed.
  • Different viewer test. Open the file in a browser and a second PDF app. Some leftovers show up in one viewer and not another.
  • Metadata test. Open the document properties (Cmd+I in Preview's Inspector) and check the author, title and keywords fields for sensitive details.
  • Attachments and comments. Check that no embedded files, comments or form fields still hold the information. Also check bookmarks and the page thumbnails sidebar.
  • File name and size. Check that the file name itself doesn't reveal the person or case. If the file size barely changed after redacting a large image, the image may still be embedded.

If you're sharing something high-stakes, such as legal, medical or financial records, ask a colleague to run the same tests on the file you plan to send. A second pair of eyes finds things you've stopped seeing.

What mistakes do people make most often?

The most common mistakes are drawing boxes without applying redactions, forgetting that repeated details appear elsewhere, and sending the wrong version of the file. All of them are procedural, so a short checklist prevents them.

  • Marking but not applying. The boxes look black, but the tool hasn't deleted anything yet.
  • Redacting once, not everywhere. A name in the body is blacked out but appears in the header, a footnote, an image caption or the table of contents.
  • Leaving context that identifies the person. Removing a name but leaving a unique job title, address fragment or rare event can still identify them. Ask whether a reader could work out the answer from what's left.
  • Redacting numbers partly. Black out all digits you don't need to show, not just some. If you want to show the last four digits of an account number, redact the rest completely.
  • Forgetting the file name. Smith-divorce-settlement.pdf gives away what the contents may hide.
  • Sending the original by mistake. Keep clearly named folders for originals and shareable copies.
  • Editing after redacting. Merging or compressing a redacted file is fine, but run the checks again on the final output.

Should you redact PDFs online or on your own computer?

For sensitive documents, doing it on your own computer is the safer default, because the unredacted file never leaves your machine. An online tool needs you to upload the original, which is the very thing you're trying to protect. Even tools that promise to delete files afterwards add a step you can't inspect. We go deeper in offline vs online file converters.

Online tools aren't automatically unsafe, and some are fine for low-risk documents such as a public flyer. But if the file contains health, legal, financial or identity details, the fewer places it travels, the better.

The order of operations matters when you combine tasks:

  • Merge first, redact second. If you're assembling a bundle from several documents, combine them first so you can redact in one pass and check one file. Our guide to merging PDFs on a Mac covers page order.
  • Redact, then compress. Compression can re-encode images and shrink the file, which is fine after redaction. See compressing a PDF on a Mac without losing readability for how to keep it legible.
  • Password-protect last. A password protects the file in transit, but it doesn't replace redaction. Once someone has the password, anything unredacted is readable.

Convert2Any is a Mac app that processes files on your computer, and its PDF tools include Redact, Merge, Compress, OCR and Passwords. If you use it or any other redaction tool, run the verification tests above on your first few files so you know how it behaves.

Frequently asked questions

Does a black box over text in a PDF remove the text?

No. A black box drawn with a shape or markup tool sits on top of the text, which remains in the file. You can usually reveal it by selecting and copying the area or searching for a word you expect to be there. Only a redaction that is applied deletes the underlying content.

Can I redact a PDF in Preview on a Mac?

Preview's markup tools, including its rectangle shapes, draw over content and shouldn't be treated as secure redaction. Whether any particular Preview version removes underlying text is not something we've established, so test the saved file with the search and copy checks above before relying on it. A dedicated redaction tool that applies removal is the safer choice.

Can redacted information be recovered later?

Not if the text and images were actually deleted and you saved a new file. It can be recovered if the redaction was only visual, if you shared an earlier version, or if the information remains in metadata, comments, attachments or bookmarks. That is why checking the final file and keeping the original private both matter.

Should I convert the redacted PDF to images to be safe?

Converting each page to an image does remove the text layer, and it can be a reasonable extra step. The drawback is that the file is no longer searchable or accessible to screen readers unless you run OCR again. Only do this after the redaction has been applied, never as a replacement for it.

How do I redact a scanned PDF?

A scan is an image, so the redaction has to alter the pixels themselves. If the scan has been made searchable with OCR, there is also a hidden text layer to remove. Check by searching for the redacted words after saving, and run OCR again afterwards if you need a searchable copy.

Explore supported formats Back to Blog