How to delete pages from a PDF
- Add your PDF. Drag it onto the box or click to browse. The tool reports the total page count, then paints a thumbnail of every page. Long documents fill in gradually, one page at a time, because each thumbnail is drawn in turn rather than all at once.
- Click the pages you want gone. A marked page turns red, dims, and picks up a cross in its corner. Click it again to change your mind. Select all and Clear cover the two extremes.
- Zoom in on anything doubtful. The magnifier button on each card opens a full-size preview around six times taller than the thumbnail. Arrow keys step through pages, Escape closes it, and there is a mark-for-deletion button inside, so you can decide without shutting the preview.
- Press Delete selected & download. The result is named after your original with
-pages-removedon the end.
Typing page numbers instead
Under the counter there is a collapsed Or type page numbers box. It takes numbers and ranges separated by commas, like 1, 3, 5-7. Numbering starts at 1, ranges include both ends, and a backwards range such as7-5 is read as 5 to 7. Overlaps and repeats sort themselves out.
The box and the grid stay in step in both directions. Type a range and those cards light up. Click cards and the box fills in with the tidy version, so picking pages 1, 2, 3 and 5 writes 1-3, 5.
Numbers always refer to the original file. Pages are stripped from the back forwards, so deleting page 2 never renumbers page 7 halfway through the job.
Your file will not get smaller
This surprises people, so here is a measurement rather than a claim. We built a three-page test PDF of 58,805 bytes where page two carried the only photograph. Deleting page two gave back a file of 58,804 bytes. One byte.
Removing a page unhooks it from the page tree, which is the index a viewer walks to find pages. The objects sitting behind that page do not go anywhere. The page dictionary, its content stream, its fonts and its images all stay put in the saved file. Write the same output with object streams switched off and you can count three/Type /Page objects in a document that now reports two pages.
That is normal for this style of editing rather than a quirk of one tool. Issue 140 in the pdf-lib project records a 13 MB, 2,000-page PDF that came back the same size after 1,900 pages were removed. Adobe ships a checkbox for the problem: the Remove Hidden Information panel in Acrobat listsdeleted or cropped contentas its own category of leftover data, alongside metadata and hidden layers.
So deleting a page takes it out of the document. It does not shred it. If the page held a salary line, a home address or a draft you would rather nobody dug up, treat the download as tidied, not sanitised. Run it throughCompress PDF or Redact PDF afterwards. Both rebuild the file page by page from rendered images, so orphaned objects never reach the output. The cost is real: text in the rebuilt file stops being selectable or searchable.
What else the edit leaves alone
Bookmarks are untouched. Nothing rewrites the outline tree, so a bookmark aimed at a page you deleted no longer lands anywhere sensible, and the same applies to any internal link that jumped to it, which is worth a quick check if you are cutting pages out of a long manual or a contract with a clause index at the front.
Digital signatures break. A signature covers the exact bytes that were signed, and you have just changed them, so a countersigned contract will need re-signing after the edit. That is correct behaviour rather than a fault.
Password-protected files get turned away at the door. Strip the password first, then come back.
When this beats splitting the file
Scanners are the usual culprit. Duplex feeders add a blank reverse side to every single-sided sheet, and batch scanners drop a separator page between documents. Twenty blanks in a forty-page scan is an ordinary Tuesday.
Beyond that: pulling a cover sheet or an internal approval page before something goes to a client, trimming a report down to the six pages a portal will accept, or dropping the duplicate scan you made when the first one came out crooked. If you want most of a document, deleting the rest is quicker than extracting. If you want a handful of pages out of two hundred, Extract PDF Pages is the better fit.
Frequently asked questions
Is there a page or file size limit?
No fixed limit. Available memory on your device is the real ceiling, and thumbnail rendering is what tends to slow down first on documents running to several hundred pages.
How do I type the page list?
Commas between entries, a dash for ranges: 1, 3, 5-7. Numbering starts at 1 and both ends of a range are included, so 5-7 covers three pages.
Why will it not let me delete every page?
A PDF has to contain at least one page to be a valid PDF. Select the lot and you get a message asking you to leave one behind instead of a file no reader will open.
Why is the download the same size as the original?
Because removing a page removes it from the page tree, not from the file. Fonts, images and content streams belonging to the deleted pages are still written out. Our three-page test file shrank by a single byte after losing the page that held the only image.
Does deleting a page really remove the data?
Not in the sense a lawyer or a compliance officer would mean. The page stops being visible and stops being printed, but its bytes remain recoverable from the file. For genuine removal, rebuild the document with Compress PDF or Redact PDF.
What happens if I type a page number the PDF does not have?
Nothing happens. The parser checks your list against the total page count, rejects anything outside it, and leaves your current selection exactly as it was. The download always follows what is marked in the grid, so a typo cannot produce a broken file.
Related: split a PDF into parts · reorder pages · merge PDFs · rotate pages.