How to Split PDF Files by Page Range, Size or Chapter

Most people reach for a PDF splitter because a single file is doing the work of several. A thesis committee wants chapter 3 on its own, an accountant wants each customer's invoice as a separate attachment, or a sheet feeder has turned a morning of scanning into one 80-page file that mixes a lease, two pay stubs and a stack of receipts. In each case nothing on the pages is wrong. The problem is only where one document ends and the next begins, and splitting moves those boundaries without touching the content.

The fourth job is about size rather than structure. Upload forms for job applications, insurance claims and university admissions often cap each file at 2 MB, 5 MB or 10 MB, and Gmail allows 25 MB per message, attachments included. A 40-page scan at 300 DPI can easily pass 20 MB. Splitting it into parts that each fit is often better than compressing the whole file until fine print turns soft, because every page keeps its original resolution.

Page ranges are the most flexible split because you say exactly where each cut goes. The usual syntax is a comma-separated list where a hyphen joins the first and last page of a range, so 1-3, 7, 9-12 means pages 1 to 3, page 7 alone and pages 9 to 12. Some tools turn every range into its own file, giving three files here; others gather all the listed pages into one 8-page PDF. Read the field label first, because the same text produces different results.

One file per page and every N pages are the mechanical options. They need no planning, only a page count, and they suit stacks where every document has the same length. A 60-page scan of two-page forms split every 2 pages gives 30 files, and if the total does not divide evenly the last part simply holds the remainder. Splitting by file size works the other way round: the tool keeps adding pages to a part until the next page would push it over the limit, then starts a new part, so parts have different page counts but each one fits the upload form.

Bookmark splitting relies on the document outline, the clickable table of contents most viewers show in a sidebar. The PDF standard, ISO 32000, calls it the outline, and each top-level entry points to a page. A splitter can start a new file at each of those pages and name it after the entry, turning a 400-page manual into chapters without counting pages. Scans and many exported reports have no outline, and then you are back to page ranges.

You may not need a separate tool for a single cut. Every desktop operating system can print a page range to a new PDF, and macOS Preview can also drag pages out of a document. The catch is that printing re-renders the pages: the new file usually loses its bookmarks, internal links and fillable fields, and a scanned page may be recompressed. That is fine for a two-page extract you will attach to an email, less fine for a contract with form fields or a manual whose readers depend on the outline.

Adobe Acrobat is the exception among installed apps because its split command copies pages rather than printing them. It sits in the Organize Pages tool and offers splitting by number of pages, by file size and by top-level bookmarks. The free Acrobat Reader can open and print PDFs but cannot split them, so the Organize Pages route needs a paid Acrobat plan. The tabs below give the exact clicks for each route.

A browser tool is the practical route on a Chromebook, a work PC where you cannot install software, or a phone, and it does what the print dialog cannot: several cuts in one pass with the document structure kept. With the UltKit PDF splitter you upload one PDF of up to 95 MB, choose a Split method and fill in the one field that appears for it. Open the file in any viewer with a thumbnail sidebar first and note where each document starts; the page numbers you type are positions in the file, counted from 1, not the numbers printed in the page footers.

The method names map onto the five splits above. Custom ranges turns every range in the Ranges field into its own PDF, so 1-3, 7, 9-12 gives three files. Extract pages takes the same kind of list and builds one new PDF from it. Every page, Every N pages, Max size per file and By bookmarks (chapters) cover the rest, with Pages per file accepting 1 to 1,000 and Max file size accepting 1 to 100 MB.

A duplex scanner captures both sides of each sheet, so a 20-sheet stack becomes a 40-page PDF where page 1 is the front of sheet 1, page 2 its back, page 3 the front of sheet 2 and so on. That order matters when you split. If every sheet is a separate document, such as a stack of two-sided forms, split every 2 pages and each sheet becomes its own file with its back attached. If the backs are all blank, you probably want the odd pages only, which drops every empty back in one step.

Pulling odd pages is a range job. With Extract pages, a list such as 1, 3, 5, 7 builds one PDF from the fronts, and every odd page of a 100-page scan fits comfortably inside the 300-character limit. Acrobat's print dialog has a dedicated subset setting, Odd or Even Pages under More Options, which you can combine with a PDF printer when the document has no links or fields to preserve. The even pages give you the backs on their own, useful when someone wrote notes on the reverse of a printed form.

Single-sided scanners create a different problem. If you scanned all the fronts, flipped the stack and scanned the backs, the file holds the fronts followed by the backs in reverse order. Splitting will not fix that; fronts and backs have to be interleaved, which is a merge task. Split the file into its fronts half and backs half, recombine them, and check the last sheet, where reversed stacks usually go wrong.

A split that copies pages, rather than printing them, can carry the document structure into each part, but some of it depends on pages the part no longer contains. A bookmark pointing at page 45 belongs in whichever part holds page 45 and nowhere else. A link from page 3 to page 60 has nowhere to land once those pages sit in different files, so it stays visible but does nothing. Digital signatures are the strictest case: a signature covers the bytes of the whole signed file, so a part cut from a signed contract will not validate as signed.

File names are the part people skip and regret. Parts named split_1.pdf to split_12.pdf force the recipient to open each one to find out what it is. Name parts by page range when someone needs to trace them back to the original, by part number when order matters more than content, and by bookmark title when the parts are chapters. The list below covers each element and the setting that controls it.

Splitting is fully reversible as long as you keep the parts. Because pages are copied unchanged, merging the parts back in order gives you a file with the same pages, the same text and the same images as the original. What may differ is the outline, since each part only carried the bookmarks for its own pages. If you split a 2 MB-capped upload into four parts, the receiving office can merge the PDF parts back into one file without any loss, and our guide on how to merge PDF files covers ordering and bookmarks in detail.

Keep the original until you have opened every part. It is the only copy that still holds the full outline, cross-part links and any digital signature, and it is far easier to split again from the original than to rebuild it from twelve pieces. If a part is still too big for its portal after splitting, run that one part through compress the PDF rather than the whole document, so the other parts keep their original image quality.

Split jobs often involve exactly the files you would least like to leak: bank statements, medical records and contracts. On UltKit, files are deleted 30 minutes after processing, there is no account to create, and nothing is installed. For records you may not share with any third party, use Preview or Acrobat locally and accept the limits described above.

Most failed splits come down to four causes: a password on the file, page numbers that do not match the file's own count, blank pages from the scanner, or a page that is larger than the size you asked for. Each has a direct fix, and none of them require starting the scan again.