How to Extract Pages from a PDF Into a New File

Extracting pages means copying a range out of a PDF into a new file, leaving the original alone. It is one of the few PDF operations that is genuinely lossless and genuinely simple — the pages are copied as they are, with their text, fonts and images intact.

Extract, split, delete — three different things

Worth separating, because tools name them inconsistently:

Extract copies selected pages into a new file. The original is untouched.

Split divides one file into several — commonly every N pages, or at bookmarks.

Delete removes pages from the original. Destructive, and the operation people reach for when they meant extract.

If you want pages 5 to 12 to send to someone, you want extract. If you want to be rid of pages 5 to 12, you want delete. Choosing the wrong one on your only copy is a bad afternoon.

Doing it

The extract pages tool here takes a file, a page range, and gives you a new PDF containing only those pages. The original is not modified. There is also a delete pages tool for the other direction.

In Adobe Acrobat, the Organize Pages tool has Extract, with an option to leave the original intact or remove the extracted pages from it — read that checkbox before clicking.

On macOS, Preview does this without any extra software: open the thumbnail sidebar, select the pages you want with Command-click, and either drag them to the desktop or use Print with "Selected pages" and Save as PDF.

Page numbers are the usual trap

The page number printed on the page and the page number your viewer counts are frequently different. A report with a cover, a contents page and roman-numbered front matter can easily have printed page 1 sitting at viewer position 7.

Extraction tools work on viewer position — the nth page of the file. Always. So check which you are looking at before entering a range, and open the result to confirm you got what you meant.

What survives, and what quietly does not

Survives: text, fonts, images, vector content, and the visual layout. An extracted page is byte-for-byte the same page.

Often does not:

• Bookmarks pointing at pages outside the extracted range. Some tools rewrite them, some drop them entirely.
• Internal links to pages you did not extract, which become dead.
• Form fields, sometimes — particularly where fields were part of a document-wide structure.
• Digital signatures. Always. A signature covers the whole document, so any extract is an unsigned new file.
• Tag structure, sometimes partially, which matters if the document has accessibility requirements.

For a plain text-and-images document none of this applies. For a signed contract, a tagged report or a fillable form, open the result and check the thing you care about.

Extracting is not redacting

If the reason you are extracting pages is that the other pages are confidential, extraction is the right move — the removed pages genuinely are not in the new file.

What is not safe is extracting a page that itself contains sensitive content you have covered with a black box. The content under the box is still in the file and can be copied straight out. That needs proper redaction, which removes the underlying content rather than hiding it.

When you need to change the pages, not just move them

Extraction copies pages as they are. If the extracted pages also need a correction — a name, a date, a figure — that is an edit. This site's editor changes text in place with the layout preserved, on files up to 10 MB and 100 pages. Extract first, then edit the smaller file: it is faster and there is less to check afterwards.

Frequently asked questions

Does extracting change the original file? No. It copies pages into a new file. Deleting pages is the destructive operation.

Can I extract non-consecutive pages? Yes — most tools accept a list such as 1, 4, 7-9, and produce one file containing those pages in order.

Why does my extracted file have the wrong pages? Almost always printed page numbers versus viewer positions. Tools count viewer positions.

Is the extracted file smaller? Usually, though not always proportionally — embedded fonts and shared resources may be carried over whole.

Do signatures survive extraction? No. A signature covers the entire document; any extract is a new, unsigned file.