How to Redact a PDF Properly (Why Black Boxes Don't Work and How to Check Yours)

You need to send a bank statement to a landlord, so you black out the account number in a PDF editor. On screen it looks perfect: a clean black bar where the number used to be. You attach it to an email and send it. The person who receives it clicks and drags across the black bar, copies, pastes into a note, and reads your account number.

Nothing went wrong with the software. The black bar was never a deletion. It was a shape drawn on top of the text, and the text was still in the file the whole time. This guide explains why that happens, how to redact a PDF so nothing is left underneath, and a short test you can run to prove it worked before you send anything.

The Short Version

• Covering is not removing. Black rectangles, black highlights, white boxes and shapes drawn in Word all leave the original text in the file.
• Real redaction does one of two things. It either deletes the content from the file, or it turns each page into a picture so there is no text left to find.
• Find everything first, then test the result. A redaction is only as complete as the areas you covered, and you should always check the exported file before sharing it.
• If you control the source document, edit that instead. Deleting the information in the original and exporting a fresh PDF is often the cleanest option.

Why Black Boxes Fail

A PDF is not a photograph of a page. It stores text as text, in a layer that is separate from how the page looks on screen. When you draw a black rectangle in a PDF editor, you add a new object on top. The viewer paints it over the words, so your eyes see black. The words themselves are still stored underneath, which means anyone can select them, copy them, search for them, or move the box away to see what it was hiding.

The easiest way to picture it: a sticky note placed over a printed word hides it from view, but the word is still on the paper. Real redaction is cutting the word out of the page.

Real Cases

This mistake is common enough that it has produced well-documented incidents. In a January 2019 court filing in the Paul Manafort case, passages were covered with black bars, and reporters were able to select the bars, copy the text, and paste it into a new document to read what was underneath. A 2009 airport screening manual from the US Transportation Security Administration was published online with black boxes over its sensitive sections and had the same weakness. In both cases the people involved could see black bars and reasonably assumed the content was gone.

The Ways People Get Redaction Wrong

The Two Meanings of "Flatten"

You will see advice online that says flattening a PDF does not fix a redaction, and other advice that says flattening is the solution. Both can be right, because the word is used for two different things.

Flattening annotations merges drawn objects, such as your black box, into the page content. The box becomes part of the page, but the text beneath it is still stored as text. This does not redact anything.

Flattening the whole page to an image renders the page, black box included, into a picture and builds a new PDF from those pictures. Because the new file contains only images, the original text and vector content no longer exist in it. This does remove what is underneath, and it is the method used by the ToolsCrow PDF Redaction tool.

Two Methods That Genuinely Work

1. Remove the content from the file. Professional PDF editors such as Adobe Acrobat Pro have a true redaction feature: you mark the text or images to remove, then choose Apply, and the content is deleted from the file rather than covered. The rest of the document stays searchable, which matters for legal work and accessibility. The drawback is that this is typically paid desktop software, and you have to use the real redaction command, not the rectangle tool sitting next to it.

2. Turn every page into an image. You draw boxes, then export a new PDF where each page is a picture. There is simply no text left to extract. This is simple and robust, and it is what the free tool below does. The trade-off is that the whole document stops being searchable.

How to Redact a PDF With the ToolsCrow Tool

The PDF Redaction tool works on a page-by-page basis, in your browser.

1. Open your PDF and move through it with Prev and Next to review every page.
2. Click and drag to draw a solid black box over each sensitive area. Use Undo Box to remove the last box on the current page, or Clear Page to remove all boxes on that page.
3. Check every page before exporting.
4. Choose Export Redacted PDF and download the new file.
5. Test the exported copy using the checklist later in this guide before you share it.

On export, each page is rendered with your boxes on it and rebuilt as an image-based PDF, so nothing from the original page survives underneath. Your original file is not changed. The tool makes a separate download, so keep the source private and send only the redacted copy. The page states that processing happens locally in your browser.

What You Give Up

Be aware of these trade-offs before you export:

• The whole document stops being searchable. Every page becomes an image, so you cannot select, copy or search any of the text, including the parts you did not redact.
• The file can be larger. Page images usually take more space than text, so the output may be bigger than the source.
• Screen readers cannot read it. Text that has become an image is not available to assistive technology.
• It does not find sensitive data for you. The tool covers only the areas you draw, and it does not check legal compliance. For high-stakes disclosure, follow your organization's or court's review process.

Before You Redact: Find Everything

The most common real-world failure after "using black boxes" is simply missing something. The same name appears redacted on page 2 and visible on page 9. A phone number sits in a footer. Spend a few minutes finding every instance before you start drawing.

1. Search the original. If the PDF is a normal text document, press Ctrl+F (Cmd+F on a Mac) and search for each name, number and address you want gone. Note every page where it appears.
2. Scanned documents are different. A scanned PDF is a picture of paper, so search will find nothing. You will need to read every page yourself.
3. Check the usual hiding places. Headers, footers, page numbers that include a case or account number, captions, signatures, comments and annotations, and repeated mentions.
4. Optionally scan the text for patterns. For a text PDF, select all, copy the text, and paste it into the Sensitive Data Detector. It highlights email addresses, phone numbers, card numbers that pass the standard checksum, US Social Security number formats, IP addresses, and API keys, and it runs in your browser without uploading the text.

Keep the Detector's limits in mind. It reads pasted text only, so it cannot open a PDF directly and cannot help with scanned pages. It looks for specific patterns, so it will not flag names, street addresses, or dates of birth. And a clean scan does not prove a document is safe. Treat it as a quick helper alongside your own reading, not a replacement for it.

After You Redact: A Six-Step Test

Never share a redacted PDF without testing the exported file, not the version you were editing. Run these checks on the copy you are about to send.

Step 4 catches more mistakes than people expect. A box drawn slightly too small can leave the bottom of a "g" or the top of a "t" visible, which is sometimes enough to guess a name. Draw boxes with a margin around the text rather than hugging it.

Step 5 matters because document properties are separate from the pages. The tool's page explains what happens to the page content, but it does not say what happens to the properties of the original file, so check them yourself rather than assuming. A report titled with a client's name in its properties can give away exactly what you hid on the page.

When Redaction Is the Wrong Tool

When You Control the Source Document

If you still have the original Word, Excel, or Google document, the cleanest approach is often to delete the sensitive information there and export a brand-new PDF. There is nothing to hide because there is nothing left. Just remember that source files can carry their own hidden extras, such as tracked changes, comments, and revision history, so clear those before exporting. Word, for example, has a Document Inspector for this.

When the Reader Needs to Search the Document

Some documents need to stay searchable and accessible after redaction, such as legal filings or documents that must work with screen readers. A page-to-image method makes the whole file unsearchable, so a true redaction tool that deletes content while keeping the rest of the text live may suit those cases better. You can also run an OCR tool on the redacted PDF afterwards to bring back searchable text for the visible parts. The redacted areas stay black because there is no text there to recognize, but check the OCR result carefully, since it can misread words.

When the Recipient Needs the Information

Before covering something, ask whether the recipient actually needs it. Some processes, such as identity checks, require the full document to be visible. Redacting too much can get a submission rejected, so confirm what is required first.

Common Situations

Frequently Asked Questions

Can someone recover the text under a black box?

If the box was drawn over the text, yes, often with nothing more than copy and paste. If the page was rendered to an image and the PDF rebuilt from that image, no text exists under the box to recover.

Is a black highlight the same as redaction?

No. A black highlight is formatting applied to text that still exists. Select the area and copy it, and the words usually come out.

Does flattening a PDF remove the underlying text?

It depends on what is meant by flattening. Merging annotations into the page does not remove text. Rendering the whole page into an image does, because the new file has no text layer.

Will my redacted PDF still be searchable?

Not with this tool. Every page becomes an image, so none of the text can be searched, selected, or copied, including the parts you left visible.

Why is my redacted PDF bigger than the original?

Page images generally need more storage than text and simple vector content. The tool notes that the output may be larger than the source.

Can I redact a scanned PDF?

Yes. The process is the same: draw boxes and export. Some scanned PDFs also carry an invisible text layer created by OCR, and because the exported file is built only from page images, that hidden layer does not carry over either.

Does the tool find sensitive information automatically?

No. It covers only the areas you draw. To help find pattern-based items such as emails, phone numbers, or card numbers in a text document, you can paste the text into the Sensitive Data Detector, which does not detect names or addresses.

Is it safe to use an online tool for a sensitive PDF?

It depends on whether the tool uploads your file. Many online tools do. The ToolsCrow PDF Redaction page states that processing happens locally in your browser. Whatever tool you use, check how it handles your file before opening anything sensitive.

Does the original file change?

No. The tool creates a separate redacted download. Keep the original private and send only the redacted copy.

The Bottom Line

A black bar on a PDF is a picture of a deletion, not a deletion. Anything that only covers text, whether a rectangle, a highlight, a white box, or a shape drawn in Word, leaves the words in the file for anyone to copy. Real redaction removes the content or turns the page into an image, and then you prove it worked by trying to select, search, and copy from the exported file.

Find every instance before you start, draw boxes with a margin, check the file properties and the file name, and send only the redacted copy.

Redact a PDF now, free, in your browser:
PDF Redaction Tool: draw boxes and export a flattened PDF
Sensitive Data Detector: scan pasted text for emails, phone numbers, card numbers and more
All PDF Tools
All File Tools