Redact PDF

A black box is not a redaction

This is the single most important thing on this page, and it is a general truth about PDFs rather than a quirk of this tool.

A PDF is a list of objects. Text is stored as text, at a position on the page. Drawing a black rectangle on top adds another object — it does not touch the text that is now behind it. The words are still there, in the file, in full, at the same coordinates.

So the document looks redacted and is not. Anyone can recover what is underneath by selecting the text and copying it, opening the file in software that lists its contents, running a command-line text extractor, or using one of the free "unredact" tools that exist specifically for this — because this mistake is common enough that a small industry has grown up around exploiting it.

It has caught out law firms, government departments and news organisations, in cases that became public precisely because someone pressed Ctrl+A and Ctrl+C on a court filing.

Do not use this tool to hide anything confidential. Not a name, an address, a bank detail, a medical fact, a salary, a signature, a case number or a client's information. The result will look convincing and will not protect it.

Check any redacted PDF in ten seconds

Worth doing on every redacted document you receive or produce, whatever tool made it. Two methods:

The copy test. Open the PDF, select all the text on the page — Ctrl+A, or drag across the black box — then copy it and paste it into any text editor. If the supposedly hidden words appear in what you pasted, the redaction is cosmetic.

The search test. Use the PDF reader's own find function and search for a word you believe is hidden. If it finds and highlights it under the box, it is still there.

Either test takes seconds, and either is enough to tell you whether a document is safe to send. Make it a habit before sending any redacted file — including files redacted by expensive software, because misconfiguration is common.

How to redact properly, using this site

There is a method that genuinely works, and it uses two other tools here.

Step one: draw your black boxes with this tool and download the result. Visually it is now correct.

Step two: flatten it. Run that file through PDF to JPG so every page becomes a picture, then JPG to PDF to reassemble those pictures into a PDF.

Why this works: converting a page to an image discards the object list entirely. The text stops being text and becomes coloured pixels — and the pixels where your black box sits are simply black. There is nothing underneath because there is no longer any "underneath". Run the copy test on the result and you will get nothing at all.

What it costs: the whole document stops being selectable and searchable, a screen reader can no longer read it, links stop working, and the file is usually larger. That is a real trade, and for a genuinely sensitive document it is the right one — this is essentially what professional redaction workflows do, which is why redacted documents released publicly are so often unsearchable.

One caution: use high quality when converting to images, or small print in the rest of the document becomes hard to read.

And if the document is legally sensitive — litigation, a regulatory filing, a data subject request, anything with a compliance obligation — use software built for it. Adobe Acrobat's redaction tool genuinely deletes the underlying content, and it also strips metadata, which is the other place information hides. That is worth paying for when the stakes justify it.

The metadata nobody checks

Worth knowing, because it defeats even a correct redaction.

A PDF carries information beyond the visible page: the author name, the software that made it, the original filename, creation and modification dates, and often revision history or comments. A document can be perfectly redacted on the page and still identify its author, or reveal that an earlier draft existed.

For anything sensitive, check and clear the document properties before sending, and be aware that flattening to images does not remove metadata — it only removes the page content. Proper redaction software handles both.

What this tool is genuinely for

Given all of the above, there is still a legitimate use, and it is a visual one.

Marking up a document for discussion. Blacking out sections to show what you propose removing, or what is not relevant to a conversation, where nobody is trying to hide anything and everyone has the original anyway.

Producing an illustration. A screenshot or example document for a presentation or a guide, where the boxes are there to keep a reader's attention on the relevant part.

Tidying your own copy of a document you are working from.

The tool itself is capable: draw boxes by hand at any position, or switch to text mode, type a word and have every occurrence boxed automatically across the pages you choose. That text mode is convenient — but note it is also the mode most likely to be used for exactly the job this page warns against, since "black out this name everywhere" is a confidentiality task, and every one of those names remains extractable.

The file is processed in your browser and never uploaded, which is a genuine privacy benefit — but it is a different thing from the content being removed, and it is worth keeping those two claims apart.

Related tools

PDF to JPG and JPG to PDF are the two halves of the flattening method above, which is the way to make a redaction real. Extract Text from PDF is how you verify it — if your redacted words show up there, they are still in the file. Crop PDF removes a region from view, and its page is clear that cropping also hides rather than deletes.