PDF Tools
Delete pages from a PDF, or pull out just the ones you need. Pick them visually, invert the selection in one click, and the original is never changed.
Overview
This free tool takes pages out of a PDF, or pulls out just the ones you want, and saves the result as a new document. Open a file, click the pages on screen, and choose whether that selection is what you keep or what you drop.
The removing half is the point. Deleting pages is the job most people actually arrive with — three blank sheets a document feeder inserted, an internal cover page that should not go outside, a section that is not relevant to this recipient — and doing it in a keep-only tool means working out the complement in your head. Selecting the pages you want gone is the whole interaction.
It produces one document, not several. If you want a separate file per page, or a long scan cut into fixed-size chunks, Split PDF is the tool for that and this one will feel like the wrong shape. The two are complements rather than alternatives.
Your original is never modified. Nothing here writes to the file you opened — it always creates a new PDF and leaves the source exactly as it was, which is why it is safe to try. Nothing is uploaded either; the document is opened and rewritten by your own browser.
Step by step
Its pages appear as thumbnails so you can see what you are choosing.
Click the pages you want to keep, or switch to remove and click the ones to drop — the tool works out the rest.
Check the page count shown before downloading. Your original file is never modified.
Background
A PDF is a set of page objects with an index saying what order they come in. Removing pages means writing a new document containing only the pages you kept, with a fresh index. The pages that survive are copied across unchanged — their text, fonts, images and vector instructions are the same objects they always were — so nothing is re-rendered and nothing loses quality.
The pages you dropped are genuinely absent rather than hidden. Their content is not copied into the new file, so it cannot be recovered from it by any amount of inspection. That makes page removal a real way to withhold whole pages before sending a document, which is worth knowing because most ways of 'hiding' content in a PDF are not real at all.
It does nothing, however, for sensitive content on a page you are keeping. Drawing a black rectangle over a paragraph in an editor leaves the text underneath, selectable to anyone who tries. Removing the whole page works; obscuring part of one needs proper redaction, which removes the underlying content rather than covering it.
The most common surprise afterwards is numbering. Numbers printed on a page are part of the page image, so they do not renumber themselves. Remove pages four and five of a ten-page report and the sheet that was six still reads 'six' while sitting fourth in the file. Any contents page still points at the original numbering, and internal links to removed pages no longer have a target.
What is lost along with the pages is anything the document held as a whole. The bookmark outline belongs to the document rather than to a page, so it does not survive; form fields defined at document level may not either; and a digital signature is invalidated by any change to the file it signed — correctly, because the document it signed no longer exists.
Reference
Nothing here is destructive to your original, so the risk is not losing the document — it is producing one that is subtly wrong and not noticing until a recipient does.
Which pages you are selecting
The thumbnails are there so this is a decision made by looking rather than from memory. Clicking a page fills in the page box, so a selection started visually can be finished by typing on a long document.
Required
Whether you are keeping or removing
The same selection means opposite things in the two modes, and the panel states the outcome as a page count before anything downloads. Reading 'keeps 3 of 12' against 'keeps 9 of 12' is the fastest check there is.
Required
Printed page numbers
They do not update, because they are part of the page. A trimmed report reads as though pages are missing, which is exactly what happened — worth a sentence to the recipient rather than a surprise.
Optional
Contents pages and cross-references
A contents page in a trimmed document points at numbering that no longer describes the file, and links to removed pages have nothing to link to. Remove the contents page too, or say which pages the extract came from.
Optional
The bookmark outline
Stored for the document as a whole, so the new file starts without one. For a long extract where navigation matters, plan to rebuild it in an editor or keep the full document available alongside.
Optional
Digital signatures
Any change invalidates them, which is the point of a signature rather than a defect. Where one has to stay verifiable, send the signed file whole and trim a copy for anything else.
Optional
Whether a page contains something worse than it looks
Removing a page removes everything on it. Keeping a page and hoping nobody reads the footer does not. If the concern is a paragraph rather than a sheet, this is not the tool.
Optional
The file name
Output is named after the source with a suffix, which distinguishes it in a downloads folder and tells a recipient nothing. Rename it before sending.
Optional
Who it helps
Document feeders insert blank separator sheets and duplicate pages that jammed and were re-fed. Dropping them is a two-minute job that turns a messy scan into a filed document.
Internal cover sheets, approval pages and pricing appendices routinely travel further than intended. Removing them before sending is far easier than explaining afterwards why the recipient can see them.
A statement or a report where only some pages are relevant to a client. Sending the whole thing because trimming looked like work is how far more information gets disclosed than anyone intended.
Disclosure means sending precisely the pages in scope and no others. Being able to see each page before it goes, and to confirm the resulting count, is a meaningful control rather than a convenience.
One chapter from a scanned book, one section from a long paper, a submission with the marking rubric taken out. The kept pages retain their original typesetting exactly, which matters for equations and figures.
Portals frequently cap file size with no flexibility. Removing pages that are not needed is often quicker than compressing, and it loses nothing on the pages that remain.
Do it properly
Noting that pages 4, 9 and 12 should go and then removing 4, 9 and 13 is an easy mistake and an invisible one. The thumbnails make it a visual check, which is both quicker and much harder to get wrong.
Expressing 'drop these three' as 'keep the other forty-seven' means doing arithmetic that the tool can do. It is also where off-by-one errors come from, because the complement of a range is easy to get slightly wrong.
The panel says how many pages the file will have. Comparing that with what you expected catches a wrong mode or a mistyped range before it becomes a document somebody has already received.
Printed numbers do not renumber, so a trimmed document looks like it is missing pages. One line — 'pages 12 to 14 of the annual report' — prevents the recipient wondering whether they were sent a fragment.
A contents page describing a document you have trimmed is worse than none: its numbering no longer matches and its links point nowhere. Either take it out or rebuild it.
Nothing here modifies your source file, so there is no reason to delete it early. Open the new document, confirm the pages, then archive — a page dropped by a mistyped range is only recoverable while you still have the original.
Removing a page genuinely deletes its content; covering text on a page you keep does not. If the sensitive material is a paragraph rather than a sheet, this tool is the wrong one and a black rectangle is worse than useless.
Avoid these
Uploading a confidential document to an online page remover
The documents most often trimmed — statements, contracts, medical letters — are exactly the ones that should not reach an unknown server. Use a tool that works in your browser, or desktop software.
Removing a page to hide something on a page you are keeping
Page removal only deletes what is on the pages you drop. Sensitive text on a retained page needs proper redaction, which removes the underlying content — a black box drawn over it in an editor leaves it selectable.
Working out the complement by hand
Switch to remove and select the pages you want gone. Typing '1-20, 24-' to drop three pages is arithmetic the tool will do correctly and you may not, particularly at the boundaries.
Expecting the page numbers to update
They are printed on the page and stay as they were. A trimmed document will look like it is missing pages, because it is — say which pages of what it came from when you send it.
Leaving a contents page in a trimmed document
Remove it, or accept that it is now misleading. Its numbering describes a document that no longer exists and its links point at pages that are not in the file.
Trimming a signed document and sending it as an original
Any change invalidates the signature, so a recipient checking it sees a failure. Send the signed file intact and trim a copy if you need a shorter version for something else.
Reaching for this when you wanted separate files
Use Split PDF. This tool always produces one document; if you need a file per page or a long scan cut into chunks, the splitter is built for exactly that and will be much faster.
FAQ
The Extract PDF Pages is at the top of this page — free, no signup, nothing uploaded.
Guides
Articles that go deeper than this page has room for.