Split PDF Pages in Half
A scanned book spread holds two pages on one sheet, and a 2-up handout holds two slides. This cuts every page down the middle — vertically or horizontally — and emits the halves in reading order, without rasterising anything.
Drop PDF files here or click to select
One or many · each page becomes two
No PDF handy? The sample library has multi-page files to try.
How the split is done
Each source page is embedded twice into a new document, each time with a bounding box covering one half, and each embedded half is drawn onto its own page at half the original size. Nothing is rasterised: text stays text, vector graphics stay vector, and the file stays selectable and searchable.
That is the important difference from cropping. A crop only changes the displayed window and leaves the whole spread in the file; this genuinely produces two pages, which is what a reader, an e-reader, or a print shop needs.
Options that matter in practice
- Right-to-left order for Japanese, Arabic and Hebrew books, where the right half of a spread is the earlier page.
- Overlap adds a sliver of the other half to each side. A percent or two rescues text that sits in the gutter of a tight book scan and would otherwise be clipped at the seam.
- Only pages wider than tall keeps portrait pages whole. Book scans usually start with a single-page cover; without this option that cover gets cut in half too.
After splitting
A book scan usually wants two more passes: auto-crop to remove the scanner margins now that each page is separate, and OCR if the scan has no text layer. If the halves came out in the wrong order, reorder pages fixes it, and rotate handles a spread that was scanned sideways.
How to split PDF pages in half
- Drop the PDF of the book scan or 2-up handout.
- Choose the cut axis — vertically for a book spread (left page | right page), horizontally for a 2-up handout stacked top and bottom.
- Let it find the gutter, or force the middle. Auto-detect scans each page for the whitest column near the centre, so a spread that was not scanned perfectly straight still splits between the pages instead of through the text.
- Set the reading order and overlap. Right-to-left for Japanese, Arabic and Hebrew books; 1–3% overlap rescues letters pinned in the gutter.
- Click Split pages and download the file — every page is now two, and the text is still selectable.
FAQ
Is the text still selectable afterwards?
Yes. The pages are embedded and clipped, not rasterised, so text, links and vectors survive intact.
My book scan loses letters at the seam.
Set the overlap to 1–3%. Each half then extends slightly past the middle, which recovers text pinned in the gutter.
How do I handle a cover page that should not be split?
Set Which pages to only those wider than tall. A portrait cover is then passed through whole while the landscape spreads are split.
Can I split into more than two?
Not here — this is a half-split. For a 4-up handout, run it twice: once vertically, then once horizontally.
Is my file uploaded?
No. pdf-lib runs in your browser and the new document is assembled locally.
Related tools
- Crop PDF — change the visible window instead of cutting
- Auto-crop margins — the natural next step for scans
- N-up imposition — the opposite: many pages onto one sheet
- Booklet imposition — reorder pages for saddle-stitch printing
- Rotate pages — fix a sideways scan