How to Split a PDF: Extract, Remove, or Split Into Parts

People say "split the PDF" to mean at least three different operations. Choosing the wrong one is the main reason this simple task takes longer than it should.

Three operations, three purposes

Extract

Take specific pages out into a new file, leaving the original alone. You want pages 5 to 12 of a 200-page report as their own document. The original is untouched, and you end up with one new file containing exactly what you asked for.

This is what most people actually want when they say split.

Remove

Delete specific pages and keep the rest as one document. You have a 40-page scan with three blank pages and a duplicated cover. You are not producing a new document — you are cleaning this one.

The distinction from extract is what you keep. Extract keeps the selection; remove keeps everything else.

Split into parts

Cut one document into several, either at fixed intervals or at specified break points. A 120-page catalogue becomes twelve 10-page files. A combined bank statement PDF becomes one file per month.

This is the least common of the three and the one people reach for by default, which is why they end up with twelve files when they wanted one.

Picking the right one

One question sorts it: how many files do you want at the end, and what is in them?

• One new file with a specific selection → extract
• One file with unwanted pages gone → remove
• Several files covering the whole document → split into parts

The common mistake is using split-into-parts and then deleting the pieces you did not want. That works, but it is three steps where one would do, and you have to clean up the extra files.

Page ranges without off-by-one errors

Nearly every PDF tool numbers pages from 1 and treats ranges as inclusive. So 5-12 gives you eight pages: 5, 6, 7, 8, 9, 10, 11, 12.

Two things to watch. First, tools generally count physical pages, not printed page numbers. If your document has four pages of front matter numbered i to iv, physical page 5 is probably printed page 1. Always work from what the viewer's page counter says, not the number printed on the paper.

Second, most tools accept comma-separated ranges — 1-3, 7, 15-20 — and extract them into a single file in that order. Useful for pulling scattered sections out in one pass.

Why splitting never costs quality

Worth stating plainly because people worry about it. A split copies page objects verbatim into a new document. Text stays as text drawing instructions, images keep their exact original encoding and resolution. No re-compression happens.

If a "split" ever produces visibly worse pages, the tool converted your pages to images somewhere in the process. That is a broken implementation, not a property of the operation.

The size surprise

Here is the counter-intuitive bit. Extract one page from a 200-page document and the result can be several megabytes.

The reason is that a page carries its dependencies. If it uses an embedded font, the font subset travels with it. If it references a high-resolution image, that image comes too. In the full document those resources were shared across many pages, so the per-page cost looked small. Extracted alone, that one page bears the whole weight.

Run a compression pass on extracted pages if size matters. It will usually cut them down substantially, because a compressor can drop the parts of the font subset that single page never uses.

What gets lost

Same document-level problem as merging, in reverse:

• Bookmarks generally do not survive. The outline belongs to the document, not the pages.
• Internal links pointing to pages outside your selection become dead. A link to page 150 in a file that now ends at page 12 has nowhere to go.
• Form fields can end up orphaned if the form's structure spans pages you split apart.
• Metadata is usually replaced rather than inherited.

For a plain text document none of this matters. For a structured report with heavy navigation, check the result before you distribute it.

Doing it privately

Splitting is one of the operations where privacy genuinely matters, because the reason you are splitting is often that the document contains something not everyone should see — one client's section out of a combined report, one employee's payslip out of a batch run.

Our PDF Tools handle extract, remove and split entirely in your browser. The file never leaves your device, so a document you are splitting precisely because it is sensitive does not get uploaded to a server to do it. If you also need to shrink the result, our guide on compressing PDFs without losing quality covers the follow-up step.

A quick checklist

• Decide which of the three operations you actually need.
• Use physical page numbers, not printed ones.
• Remember ranges are inclusive at both ends.
• Expect extracted pages to be larger than you assumed; compress if needed.
• Check bookmarks and internal links if the document relies on them.
• Decrypt first if the file is password protected.