How to Delete PDF Pages Without Mistakes

Removing pages from a PDF takes a few seconds and has an expensive failure mode: nobody notices the wrong page went until someone reads the file properly, and by then it has usually been sent. The operation is not the hard part. Being certain you are subtracting the right pages is.

The printed number is not the page number

This is the single most common way it goes wrong. Ask for pages 12 to 14 and you are addressing the twelfth, thirteenth and fourteenth pages of the file — not the pages with 12, 13 and 14 printed at the bottom of them.

Those are the same thing only in a document with no front matter. A cover sheet, a title page, a contents list, roman-numeral preliminaries, an inserted blank: each one pushes the printed numbering out of step with the file’s own count. Scans add their own offset whenever a separator sheet went through the feeder. A report whose body starts on printed page 1 can easily have it as the fifth page of the file, and every range you type afterwards is wrong by four.

So establish the offset once before typing anything. Find the page carrying a printed 1, note which position it occupies in the file, and apply that difference to everything else. Then stop trusting the arithmetic and check it against what the tool actually shows you.

Delete what you can see

Typing ranges is the fast path on a long document and also the risky one, because a number is not evidence. Clicking is slower and very hard to get wrong: Remove Pages renders every page as a thumbnail and marks the ones you click, so you are choosing pictures of pages rather than positions you believe in.

The sensible approach is both — type the range to do the work, then look at what became marked before exporting. There is one behaviour to know if you do that, because it catches people out: applying a typed range replaces whatever was already marked rather than adding to it. Click three thumbnails, then type 2,5-9 and press Apply, and those three clicks are gone; only pages 2 and 5 through 9 are marked. Type first, click second.

Use the counter as a check rather than as decoration. Decide in advance how many pages should survive — forty pages minus three appendix pages is thirty-seven — and then read the footer, which states how many are marked and how many will remain. If it says thirty-six, something extra is marked and you have found it while it is still free to fix. The range box also refuses what it cannot make sense of: a backwards range, a zero, or a page past the end of the document is rejected with the reason, so the commonest typos never reach the export at all.

What deleting a page does, and does not, do

Confirming does not strike pages out of your file. A new document is created, the pages you kept are copied into it in their original order, and that new document is what downloads. Three consequences follow from that, and they are the ones worth knowing.

Your original is untouched. It is only ever read, never modified, and the trimmed version arrives beside it as a separate download with "-pages-removed" added to the name. That is a safety net when you get it wrong, and a trap if you forget which of the two files you then attach to the email.

The pages you dropped are genuinely absent from the new file, because they were never copied into it. That is a real difference from drawing a black box over something on a page you keep, where the text sits underneath the box and can still be pulled out. It is not a general redaction method, though: it works only at whole-page granularity, and anything on a page you keep remains exactly as it was.

And because the output is a newly built document rather than an edited copy, document-level extras do not come across. It is written with Formynex as its creator and producer, so the original’s title and author fields are not carried over, and a bookmark outline is not rebuilt. On most documents nobody will ever notice. On a long reference PDF whose navigation pane people actually use, it matters — so check before you replace the original with the trimmed one.

Check the result, not just the selection

Open the download before you do anything else with it, and spend thirty seconds on four things:

1. 1The page count. It should equal exactly what the footer predicted would remain. If it does not, something was marked that you did not intend.
2. 2The pages either side of every gap you made. If you removed 12 to 14, whatever follows page 11 should be the old page 15. This is the check that catches an off-by-one, and nothing else will.
3. 3The first and last pages, which is where an accidental extra removal is least visible and most embarrassing.
4. 4Anything the document says about its own pages — a contents list, a cross-reference, a printed footer. All of it is now out of date.

That last point is the one people forget. Removing pages renumbers everything after the gap, and nothing inside the document updates itself: numbers stamped onto the pages stay as they were, and a contents list still points at the old positions. Page Numbers will stamp a fresh, correct sequence onto the trimmed document. A contents list has to be corrected wherever the document was originally written.

Removing, extracting or splitting

Three tools overlap here, and which one is right depends on which side of the job is smaller.

• Removing suits dropping a few pages from a document you otherwise want whole. You name what goes, and everything else survives — Remove Pages.
• Extracting suits pulling a handful of pages out of many. You name what to keep, which is far less error-prone when the keepers are the minority, and Extract Pages can also give you one file per page rather than one document.
• Splitting suits a rule applied across the whole document — into ranges, into equal chunks, or into parts that each stay under a size limit. How to split a PDF sets out the difference.
• If pages also need reordering or turning, do it all in one pass with Organize PDF instead of exporting three times.

One last word on size, since shrinking the file is often why the pages are going. Dropping pages does reduce it, but not in proportion — resources shared across the document, embedded fonts especially, are carried over once regardless of how many pages are left. Removing pages is usually the biggest single win and rarely the whole answer: getting a PDF small enough to email covers what to do after it.

And if some of the pages you are keeping are facing the wrong way, why a rotated PDF opens sideways again explains which of those pages are actually wrong and which are wide on purpose.