"Split" and "extract" are often used interchangeably, but they answer different needs: the first cuts one file into several, the second pulls a few pages out of the middle.
1. Three common needs, three approaches
- One file per page: archiving scans with a number per page so each can be cited separately. Use PDF split with "split by page" and download a ZIP.
- Split in half: turn a two-page contract into two files, or a collection into volumes. "Split in half" produces exactly two files.
- Take a few pages: an application only needs pages 3-5. This is closer to extraction -- split and delete, or delete the other pages in the original.
2. Check three things before splitting
Page order: a duplex scan can come out as 1-3-5 / 2-4-6. Once split, that is painful to untangle. Flip through two pages first.
Encryption: a PDF with an open password cannot be split. Remove the password in the source tool first -- and never strip protection from somebody else's document without permission.
Bookmarks: they are usually lost when splitting. Re-add them afterwards if they matter.
3. Merging pages back together
After extracting, you often need to combine the new pages with another file. PDF merge orders pages by upload order, so rename the files 01.pdf, 02.pdf before uploading instead of relying on the tool to guess.
4. Does splitting make files bigger?
Total size grows slightly: split a file into 20 single pages and each carries its own font and structure overhead, so the sum can be 10-30% larger than the original. If you are only archiving, keep the complete file. For single pages you need to send, run them through PDF compression.
5. Two details for scanned documents
Inconsistent page size: paper fed slightly crooked produces pages a few millimetres apart, which looks uneven when printed. Normalise page size before splitting.
Grey instead of colour: a colour scan of a text-only document often halves in size when converted to greyscale, with sharper letter edges -- which also helps OCR.