Merge, Split and Extract PDF Pages on Windows

Split, extract and merge answer different questions

Split turns one PDF into one file per page. Every page. Feed it a 40-page statement when you needed pages 1, 3 and 5 and you get 40 files and a sorting job. Extract pages is the tool that was meant: a chosen set, kept together, in one new PDF. Merge goes the other way and joins open PDFs in an order you set.

Say the outcome out loud before you click anything. “One attachment holding my covering letter and the annexure” is a merge. “A short PDF with only the three pages that were asked for” is an extraction. The two are easy to mix up in a menu and impossible to mix up in a sentence.

Check the order, not just that it opened

Confirm which PDFs are open and which pages actually belong in the result. Pick the output destination yourself, so the new file lands somewhere you will find it again and the originals stay exactly where they were.

Then open the result and read it, because a PDF that opens is not a PDF that is right. Three places are worth the attention: the first page, the last page, and whatever sits on either side of each join. Wrong order and one page too many are the two failures that get through, and neither is visible from the file icon. The preview draws the first 40 pages, so check a long merge in the saved file rather than in the panel.

Numbering and watermarks come last

Pages mode also rotates, deletes, inserts a blank A4 page, adds a watermark, numbers the pages and writes a header or footer. Page numbers take a position and a format. Watermarks lay translucent text across the page.

Do all of that after you have settled which pages belong in the file, never before. Number a document you are about to cut and you end up with a page 7 that follows page 4, which is the sort of thing a reviewer notices and you do not. If the finished file has to fit an upload cap, compression is the next step, not the first one.

Cropping hides. It does not remove.

Crop changes which part of the page you can see. The text outside that area is still in the file, and it comes back for anyone who looks. A yellow highlight or a black rectangle is an annotation drawn on top; the words underneath are untouched and still searchable.

Creasepoint has a Redact tool, and what makes it one is that the objects come out of the file and the saved file is then searched for the text they held - if any of it is still findable the operation fails rather than reporting success. What it cannot reach is a form field's value, which lives in the document's form data rather than on the page; it names the field and says so before it runs. That limit is written down on the limitations page for the same reason it is written down here. The single-revision output check is not a sanitiser either: metadata, attachments and cropped-away content all survive it. If something has to be gone before the file leaves your hands, this is not the application for that job.

Common questions