“Split a PDF in half” means two different things
You typed "split PDF in half", clicked the first result, uploaded your file, and got back something that is not what you wanted. Either you have two documents when you wanted narrower pages, or you have the same wide pages when you wanted a front half and a back half.
The phrase means two different jobs. Half the document — pages 1 to 50 in one file, 51 to 100 in another. Or half of every sheet — a scan of an open book, where one page of the PDF holds two pages of the book side by side, and you want one page each.
They need opposite tools, and neither tool will tell you that you picked the wrong one. It will just do its job on your file and hand you a result. Thirty seconds here saves the afternoon.
Why one phrase covers two jobs
To a PDF, a page is one rectangle with one lot of content drawn on it. That rectangle is the atom every page tool works in: split, extract, delete, reorder, merge — all of them move whole pages around, and none of them can see inside one.
So when a scanner photographs an open book, it sees one sheet of paper and saves one page. The fact that the sheet carries two book pages is invisible to the file. There is no line down the middle, no marker, nothing. It is a single wide page, exactly like a single wide page of anything else.
That is why a splitter cannot help with a book scan. Splitting divides the sequence of pages. The job you want divides the paper. Different axis, different tool.
Reading A: halve the document
You have a 120-page report and you want two files of 60. Nothing about the pages changes; only the boundary between them.
- Put the file into Split PDF in half. There is no form to fill in — the halves come from the page count.
- Run it. Two PDFs come back in one zip, named after your file with
-1and-2on the end.-1is the front half. - Check the join. Open the last page of
-1and the first page of-2and make sure the cut did not land in the middle of a table or a signature block.
The odd-page rule. A document with an odd number of pages has no exact middle, so the extra page goes into the first file — 11 pages become 6 and 5, not 5 and 6. The result says so in a note, so you never have to count. This matters more than it sounds: with a duplex-printed document, the front and back of one sheet must stay together, and a silently misplaced page is the kind of thing you find out at the printer.
If the halves should not be equal, this is the wrong tool and Split PDF by pages is the right one — you name the pages to cut after, and two numbers give you three files. Same tool if you want one file per page, which has its own short answer including the file-naming detail that stops page 10 sorting between 1 and 2.
Reading B: halve each sheet
You have a scanned book, a two-up handout, or an A3 spread, and every page of the PDF is twice as wide as it should be. This is not a page-count problem — it is a crop-and-interleave job, and it has a full write-up of its own, covering how to measure the crop, what a gutter that alternates sides does to the numbers, off-centre scans, right-to-left books, and the finishing pass once the pages are single.
The shape of it, for orientation: crop the original document twice — once keeping only the left half of every sheet, once keeping only the right — then feed both files into Alternate & Mix PDF so they interleave into reading order. Go to the dedicated post before you start; the exact crop measurements and the one mistake that wastes an hour are both there, not here.
Which one did you mean
Two questions settle it.
- Look at one page on screen. Does it show two pages of something? If yes, you are in reading B. If each page shows one page, you are in reading A.
- Is the page wider than it is tall? Nearly all reading-B files are landscape, because two portrait pages side by side make a landscape sheet. A portrait document almost always means reading A.
The awkward case is a two-up handout you made yourself with N-up PDF, where the pages are landscape and the content is two-up. That is reading B, and the crop recipe undoes it.
What our tools will not do
Split PDF in half divides the document, not the paper. It matches your search term and it will happily run on a book scan, giving you the first hundred double-wide sheets and the last hundred double-wide sheets. It is named here so you do not click it hopefully.
Neither route keeps bookmarks. Both halves in reading A are new documents, and so are both crops in reading B, so an outline and a form's field definitions are left behind.
When a command line is better
If you are doing reading A more than once, this is faster and we would rather you finished the job:
qpdf --pages in.pdf 1-60 -- first.pdfdoes reading A from a script, and keeps the document structure that a rebuild drops.
Reading B's command-line options — mutool poster, Briss, ScanTailor — are listed on
the dedicated post, along with the browser recipe.
Putting it back together
Both readings are reversible. Two halves of a document go back into one with Merge PDF, in the order you arrange them — useful when you split a file only to work on one half. A crop is reversible too, by cropping the original again with different numbers, which is exactly why the finished file still contains the half you cannot see.
The one thing worth doing before either job is checking what you actually have. Open a page. Count what is on it. That is the whole diagnosis.