Splitting a PDF by page ranges, explained

“Send me pages 4 to 9” sounds like a simple request until the file's page 4 has a big roman numeral iv on it, or the person meant the chapter that starts on printed page 4. Page-range notation is the short, precise way to say which pages you mean, and it is used by print dialogs, PDF tools and office software alike. Once you understand a few rules, you can cut any document exactly where you intend.

The notation

A range list is a series of items separated by commas. Each item is either a single page or a span:

Spaces do not matter, so 1 - 3 ,5 means the same as 1-3,5. PDFNova also accepts semicolons as separators and an en dash (–) in place of a hyphen, since word processors often swap one for the other automatically.

Positions, not printed numbers

This is the source of most mistakes. In range notation, page 1 is the first page in the file, whatever is printed on it. Books and reports often have front matter — a cover, a title page, a table of contents numbered i, ii, iii — so printed page 1 might be the seventh page of the PDF.

Some PDFs contain page labels, a feature that lets a reader display “iv” or “A-3” in its page box. Labels are a display convenience; they do not change the underlying order. To find a page's position, look at the thumbnails in a tool that numbers them by position (as PDFNova does), or use the “page X of Y” counter in your reader rather than the number printed in the footer.

A practical habit: when someone asks for “pages 4–9,” confirm whether they mean printed numbers. If they do, find the position of printed page 4 and add the offset. If printed page 1 is position 7, printed 4–9 is positions 10–15.

Splitting versus extracting

The same range list can mean two different things, depending on the tool.

• Extract takes every page mentioned and produces one file. 1-3, 5, 8-10 gives a single 7-page PDF.
• Split by ranges produces one file per item. The same list gives three files: pages 1–3, page 5, and pages 8–10.

Use Extract pages when you want a shorter version of the document, and Split PDF when you want separate pieces — chapters for different readers, or individual invoices out of a monthly batch.

Edge cases, and how a careful tool handles them

Reversed ranges

9-7 almost always means “pages 7 to 9” typed in a hurry. PDFNova reads it that way and tells you it did so, rather than silently producing nothing or reversing the page order.

Pages that do not exist

If a document has 10 pages and you type 8-12, should the tool quietly give you 8–10? It is tempting, but it hides mistakes — perhaps you opened the wrong file, or the document lost pages somewhere. PDFNova stops and tells you how many pages the file really has. Use 8- when you mean “to the end.”

Page zero and stray characters

Page numbering starts at 1, so 0 is rejected. So are items that are not numbers or spans, such as 1-2-3, 4a or 1.5. A clear error is better than a guess about what you meant.

Duplicates and overlaps

When extracting, 1-5, 3 mentions page 3 twice. Including it once is the sensible result, and the tool says so. When splitting, overlap can be intentional: 1-5, 5-9 makes two handouts that share page 5, and each file gets its own copy.

Patterns for common jobs

Chapters of a book or thesis

Look up the starting position of each chapter in the thumbnails and write one range per chapter: 1-12, 13-40, 41-77, 78-. Splitting by ranges gives one file per chapter, named after their page spans.

A batch of identical forms

A scanner often produces one long PDF from a stack of two-page forms. Splitting “every 2 pages” gives one file per form. If the stack contains a stray blank page, the remaining forms will be misaligned, so check the thumbnails first and delete any blanks.

Removing a cover and back matter

To keep everything except the first page and the last two of a 50-page file, extract 2-48. Or delete 1, 49-50 — the same result reached from the other direction.

Odd or even pages only

There is no special syntax for odd or even pages in most tools, including this one. Type the list explicitly (1,3,5,7,9). For long documents, generating the list in a spreadsheet and pasting it in is quicker than typing.

What splitting does not change

Splitting copies pages; it does not re-render them. Text stays selectable, images keep their quality, and fonts come along with each part. However, some things belong to the whole document rather than any one page: bookmarks, the document outline, interactive form definitions and file attachments. These are not carried into the parts. Links that point to a page that ended up in a different part will no longer lead anywhere.

You may also notice that each part is larger than its share of the original. If every page uses the same embedded font, each part must contain its own copy of that font. That is expected, not a fault.

Quick checklist

1. Confirm whether the request uses printed page numbers or positions.
2. Check the page count and the thumbnails before typing ranges.
3. Decide whether you want one file (extract) or several (split).
4. Use N- for “to the end” instead of guessing the last page number.
5. Read the summary of files to be produced before running the split.