Not every split is the same job. Breaking a scanned stack into one file per document is mechanical — every page becomes its own file, no judgment involved. Breaking a 300-page book into chapters, a court filing into exhibits, or an annual report into quarters is different: the document has meaningful breaks, and the split's whole value is putting them in the right places. That's page-range splitting — selecting exact runs of pages, each run becoming its own PDF. This guide is the deep dive on doing it well. For the overview of all three split modes and the tool's basics, the reference guide is How to Split a PDF File Into Separate Pages; everything below assumes that context and spends its words on ranges alone.
Thinking in ranges
A range is a consecutive run of pages that becomes one output file. A range split is therefore nothing but a set of decisions about boundaries: where each chunk starts, and where it ends. Write the recipe down before touching the tool — "pages 1–3, page 5, pages 7–9" — and the mechanical part takes care of itself.
Two properties are worth internalizing. Ranges select what goes in: any page you don't include in a range simply isn't in any output. And each range is its own file: three ranges, three PDFs. Together those two properties make range splitting as much a filtering operation as a dividing one — the pages you skip are left behind in the original, which is often exactly what you want.
How range selection works
Click a page thumbnail to select it, or drag across several to select a range; build as many separate ranges as the job needs — you get one PDF per range. A live preview shows exactly how many files you'll get and which pages are in each, before anything is written. When the split produces multiple files, each one is a separate PDF bundled into a single .zip for one download. And because pages are copied across exactly, with no recompression, every output is a normal, standalone PDF — the same quality as the source.
Recipe: a book into chapters
Chapters are the classic range job, and the classic way to get it wrong. The right process:
- Scroll the document and find each chapter's actual first page — recognized visually, by the pattern of a chapter opening: a big heading with white space above it.
- Note the page numbers as the grid reports them, not as the printed folios claim and not as you remember them.
- Select each chapter as one range, from its first page to the page before the next chapter starts.
- Run the last range all the way to the final page.
The error this prevents is the off-by-one: chapter three starts on page 41, not 42, and a range that begins one page late puts the previous chapter's last page at the top of chapter three's file. Two minutes of scrolling beats re-doing the split.
Recipe: a filing into exhibits
Litigation and filing bundles are exhibit after exhibit: Exhibit A is pages 1–4, Exhibit B is pages 5–11, and so on down the line. If you're assembling the bundle yourself, the oldest scanner-operator trick pays off: put a cover sheet naming each exhibit in front of it before scanning. The covers make boundaries unmistakable in the page grid, and they double as labels once the file is split.
The filing-specific mistake is the page between two ranges: selecting A's pages and C's pages while never quite deciding about a transition page that ends up in neither. This is what the preview is for — before splitting, walk the exhibit list against it and confirm every page is accounted for somewhere.
Recipe: a report into quarters
A year-long report into Q1–Q4 is ranges at their most seductive and their least necessary. If the quarters are genuinely equal-length — exactly 25 pages each in a 100-page report — the "every N pages" mode does the job with zero manual selection. Ranges earn their keep when the breaks are irregular: Q1 runs 32 pages because March was busy, Q2 runs 19. Find the real starts, set ranges around them, and let the lengths fall where they may.
The classic mistakes, in one list
- The off-by-one start. A range starting one page late or ending one page early puts content in the wrong file. The cure is scrolling to the real starts — never trusting remembered page numbers, and never trusting the printed folios, which describe a document that may since have been merged or reordered.
- The forgotten page between ranges. A page that belongs to neither range you selected belongs to no output at all. Check the preview as a checklist: every page you need should appear in exactly one file.
- Assuming ranges can overlap. The title page that should open every chapter file, the summary that belongs in every section — a range split gives you one PDF per selected range, and nothing about it promises that a page can land in two outputs at once. If you need a page in several files, plan a second pass for the shared pages rather than discovering this live.
- Splitting before the document is final. Merge, reorder, and trim first; split last. Every structural edit invalidates the page numbers your ranges were written against.
Check the preview, then commit
The preview is the last line of defense and the cheapest quality check in the whole process: it shows the number of files and the pages in each. Count the files against your recipe — six chapters, six exhibits, four quarters. Read each range's first and last page number against your notes. Only then split. The output arrives as one .zip of standalone PDFs, pages copied exactly as they exist in the source — and if a boundary was wrong after all, the original is untouched, and the fix is re-splitting the one range that moved.