How to remove metadata from a PDF before you send it

CVs, contracts, invoices, reports and court filings travel as PDFs, and each one carries a small dossier about itself: who made it, with which software, when, and sometimes what it said before the last round of edits. Most of it was written automatically by apps you used along the way, and very little of it appears when you open the document.

This guide maps where PDF metadata is stored, explains why editing the Properties dialog is not enough, and walks through a full clean, including the two cases where you must take an extra step: encrypted files and documents that will be digitally signed.

Where does a PDF keep its metadata?

A PDF is a collection of numbered objects plus a trailer that says where to find the important ones. Metadata is spread across several of them, and the Properties dialog in your reader shows only the first few rows of this table.

Two rows deserve a note. PieceInfo is where design apps such as Illustrator keep their own private data, which can be large. And the document ID is meant to identify a document across versions, so two PDFs sharing the first half of an ID can be traced back to the same original.

Why doesn’t clearing the Properties fields work?

Because the dialog edits one copy of data that is often stored twice, and because of how many editors save.

• Duplicates. The author, title and dates usually exist in both the Info dictionary and the XMP packet. Clearing one and not the other leaves the name in the file for any tool that reads XMP.
• Incremental saves. The PDF format lets an editor append changes to the end of the file instead of rewriting it. The old objects, including yesterday’s Info dictionary, can stay in the file behind the new ones.
• Fields outside the dialog. Comment author names, custom keys added by company systems and per-page XMP never appear in the Properties box at all.
• The export writes its own. If you clean a Word file with File → Info → Check for Issues → Inspect Document before exporting, the PDF still gets a fresh Creator, Producer and set of dates.

Cleaning the source document is still worth doing, because it removes tracked changes and hidden text from the content itself. But the metadata problem is solved by rewriting the finished PDF, not by editing it.

How do you remove metadata from a PDF?

1. 1Open the PDF metadata remover and drop in one PDF or several. Each is opened in your browser and its properties, XMP, comment authors, document ID and revisions are listed.
2. 2Leave the mode on Everything, which this page starts in. AI & provenance only exists too, for removing Content Credentials and AI-tool names while keeping the rest.
3. 3Press the remove button. The file is loaded with the pdf-lib library and saved fresh without the metadata. Pages, text, fonts, images, links, bookmarks, form fields and attachments are carried over as they are.
4. 4Check the result. The cleaned PDF is inspected again, and the new random document ID appears under Kept on purpose: readers expect one, and this one links to nothing.
5. 5Download it, or the whole batch as a ZIP.

Why can’t an encrypted PDF be cleaned?

An encrypted PDF stores its text strings and content streams in encrypted form. Rewriting the file means decrypting and re-encrypting them, which requires the password. This includes PDFs that open without a password but are locked against editing or printing: the restrictions are enforced by the same encryption.

Rather than produce a damaged file, or quietly hand back an unprotected one, the tool refuses encrypted PDFs with a message saying so. The fix is to open the document in the app that protected it, save an unprotected copy, clean that copy, then add protection again if the document needs it.

Should you clean a PDF before or after signing it?

Before. A digital signature covers the exact bytes of the file at the moment of signing. Removing metadata rewrites the whole file, so cleaning a signed PDF would leave a document whose signature no longer validates. The order that works: finish the content, clean the metadata, then sign it or send it out for signature. Empty signature fields are ordinary form fields and survive cleaning.

What is a safe order for sending a PDF?

1. 1In the source document, accept or reject tracked changes, delete comments you do not mean to send and run the editor’s document inspector.
2. 2Strip location and camera data from any photos before you insert them.
3. 3Export or print to PDF, and redact anything sensitive with a real redaction tool.
4. 4Run the finished PDF through the metadata remover and read the second inspection.
5. 5Sign it, if it needs signing.
6. 6Give the file a neutral name. The file name is not metadata, and no remover changes it.

Frequently asked questions

Will removing metadata change how the PDF looks?

No. Nothing is re-rendered, flattened or recompressed, so the document looks, prints and searches the same and forms stay fillable. Only metadata, app-private data, old revisions and the document ID change.

Can I see a PDF’s metadata without removing it?

Yes. The metadata viewer lists everything the remover would find, including comment authors and the number of earlier revisions, and changes nothing unless you ask it to.

Will the PDF get smaller?

Usually only slightly, since properties take up little space. Files with large app-private data or many incremental saves can shrink a lot more. Images are deliberately not recompressed.

Is my PDF uploaded?

No. The PDF is read, rewritten and checked inside your browser tab. Contracts and medical letters should not pass through a stranger’s server, and you can confirm in DevTools → Network that none of it does.