How to Extract a Chapter or Contract Pages From a PDF
Short answer: find the first and last page of the section you need, type that range into a PDF splitter, and download the result, a new file containing only those pages, copied straight from the original. The whole job takes about four minutes, and the part people get wrong is not the extraction, it is the boundaries: a chapter that starts on page 47 usually ends with a paragraph on page 69, and a contract clause that matters usually has its signature block three pages later. Everything below covers how to pin down the exact range, what happens to page numbers and links after you cut, what to send alongside the file, and the cases where pulling pages out is the wrong tool entirely.
Why send 20 pages instead of 320?
Because most of the people asking you for a PDF are asking for a section, not an archive. A landlord wants the clause you are disputing, not your whole lease. A professor wants chapter 6 for the reading, not the entire textbook scan. A court filing specifies the pages being cited; an accountant wants the schedule, not the full return; a vendor wants the statement of work, not the 90-page master agreement it sits inside. Three separate reasons point the same way:
- Attention. A recipient who opens a 320-page file at page 1 has to hunt. An extract that starts at the right page reads as if it were written for them.
- Size. Many portals, email systems and court e-filing services cap uploads; a chapter usually clears the cap where the full document does not.
- Privacy. This is the important one. Extraction is the cleanest form of redaction available to a non-expert: the pages you do not send are not blurred or hidden. They are absent. Nobody recovers a page that was never in the file.
That last point deserves emphasis, because people reach for a highlighter-blackout approach when a scissors approach exists. If a lease contains your former address on page 2 and the dispute is on pages 12–19, sending 12–19 solves the problem completely, with no tool that can undo it afterwards.
Step 1: Find the exact first and last page
Get this right and the rest is one keystroke. Two reliable methods:
- Follow the table of contents. Click the chapter or section entry in your reader's sidebar and look at the page box. It shows where you landed, usually as
47 of 320. That first number is the one you want. - Search a distinctive phrase. Press Ctrl+F (Windows) or Command+F (Mac) and type a phrase you remember from the section: the clause heading, a section number like 4.2(b), a sentence unique to that chapter. The match lands you on the exact page.
Then make the two checks that separate a usable extract from a wasted one:
- The last page matters more than the first. Chapters routinely continue with a paragraph or a figure caption on the following page. Contract clauses break across pages, and the initial block or signature page often sits several pages below the clause you are citing. Read to the end of the section before you settle the final number.
- Printed numbers and PDF numbers are not the same thing. A book scan with roman-numeral front matter, a report with a cover and a revision history, or an exhibit that restarts its own numbering will all disagree with the position shown in the reader. Trust the reader's page box. That is what a splitter counts.
Step 2: Extract the range
Open the PDF Splitter and drop the document onto the dropzone; the page count appears as soon as it loads, which gives you the ceiling for your range. Type what you found:
47-68
and add pages that live elsewhere in the document with a comma: 47-68, 301 pulls the chapter plus the signature page from the back of the contract. Pages come out in the order you type them, so you can deliberately put the signature page right after the clause it belongs to. Click Extract pages, and the result list shows each output with its page count and exact size before you download it. Tick the option to give every selected page its own file if what you actually need is one file per receipt, form or exhibit.
Two things worth knowing about what just happened: nothing was uploaded (the read, the split and the download all happen inside your browser tab, which you can watch in DevTools → Network if you are the trusting-sort-verify type) and nothing was re-encoded. Pages are copied, not photographed, which is the subject of the next section.
What happens to page numbers, links and bookmarks
An extract is a faithful copy of specific pages, which has predictable consequences:
| Element | After extraction | What to do about it |
|---|---|---|
| Footer reading “47 of 320” | Unchanged, because it is part of the page | Leave it and state the range in your message; matching the original is a feature for anyone verifying your excerpt |
| Links between pages inside the range | Still work | Nothing to do |
| Links or references pointing outside the range | Go nowhere useful | Mention it: “section 4.2 is cited on page 90, not included” |
| The document's bookmark list | Reflects the new, shorter file | Rename the file so the bookmark panel makes sense |
| Interactive form fields | Simple fields survive; calculated or signature widgets may not | Fill and flatten such forms before splitting |
On numbering, the experienced answer is: do not renumber. Recipients of legal, academic and financial documents routinely check an excerpt against the source, and an excerpt whose page 1 says “47” is easier to verify than one whose page 1 says “1”. One line of context (“pages 47–68 of the original agreement”) solves it in the reader's head instead of in the file.
What to send with the extract
The file is only half the job. A short, professional package looks like this:
- A filename that says what it is:
Lease-2026-pages-12-19.pdf,Contract-4.2-clause-pages-47-68.pdf. Notscan0001.pdf, notfinal_FINAL2.pdf. Recipients file by name, and courts, landlords and professors all notice. - One sentence of provenance: source document, range, and why those pages. “Pages 12–19 of the signed lease cover the entry clause we discussed.”
- The right size: if the section still will not fit the portal, compress the extract rather than the original: a 20-page section responds far better than a 300-page document. See how to compress a PDF without losing quality.
- An offer, not an dump: say the full document is available on request. That reads as helpful rather than withholding, and it keeps the bulk transfer a deliberate choice instead of a default.
When extraction is the wrong move
Honest limits, so you do not discover them at the worst moment:
- You need to hide something that is inside a page you are sending. Extraction removes whole pages, not parts of them. A clause quoting your account number still leaks on page 14; that needs actual redaction, not omission.
- The PDF is password-protected. It cannot be read without the password. Open it in a reader and use Print to PDF to make an unlocked copy, then split that, but only when you are entitled to open it at all.
- It is a scan with no text layer. Extraction still works perfectly (pages are copied either way), but you cannot search inside it to find the right page, so you will be scrolling. The range syntax cheat-sheet and quick buttons make the rest fast.
- The interactive form is the point of the document. Fill and flatten it first, because complex widgets can lose behaviour when a page is copied.
- Everyone genuinely needs the whole document. Sometimes 320 pages is the actual deliverable. In that case stop cutting and start compressing instead.
What we actually tested
We ran this the way the guide describes it: a multi-hundred-page document, a section found through the table of contents, a phrase search to confirm the boundary, then a mid-document range plus one out-of-order page pulled in a single run. The output was checked page against page against the original (text still selectable, fonts and layout unchanged, links on the copied pages resolving, the out-of-order page sitting exactly where the typed order put it) and the Network panel stayed empty throughout, which is the whole reason this can be run on a contract before it is signed rather than after. The extracted file's size matched what you would predict: the same per-page weight as the source, multiplied by the page count, because nothing was recompressed.
What this workflow does not do: it does not rewrite footers, it does not repair links that point outside the range, it does not defeat a password, and it does not make a page-level decision about a paragraph you wanted hidden. Those are different jobs with different tools, and knowing which one you need is most of the work.
Quick FAQ
How do I extract specific pages from a PDF?
Open the PDF Splitter, drop the document in, and type the pages you want: 47-68 for a chapter, 12-19 for a contract section, or 47-68, 301 to add one page from elsewhere. Click Extract pages and download the result. The pages are copied rather than re-rendered, so the text stays selectable and the file never leaves your device.
How do I find which pages a chapter or clause is on?
Jump to it from the table of contents and read the page position the reader shows (the box usually reads something like 47 of 320) or search for a distinctive phrase from the section with Ctrl+F on Windows or Command+F on Mac. Check the last page as carefully as the first: chapters spill onto the next page and contract clauses continue across the break, and the printed number in a footer can differ from the PDF's own page count when the document has front matter.
Will the extracted pages still say page 47 of 320?
Yes, because the footer text is part of the page content and extraction does not rewrite it. That is usually what you want: a recipient checking your excerpt against the original can match the numbers exactly. If the numbering would confuse them, say it in one line of your message (pages 47-68 of the original agreement) rather than trying to renumber the file.
Does extracting pages change the quality or the file size?
No: a split copies the page objects themselves, so nothing is re-encoded and there is no quality loss. The extracted file simply carries fewer pages, which is why it is much smaller. If the section is still too large to email, run the extract through the PDF Compressor afterwards; compressing a 20-page section is far more effective than trying to squeeze the whole 300-page document.
Can I extract pages from a password-protected PDF?
Not directly: an encrypted file cannot be read without its password. Open it in any PDF reader, use Print to PDF to produce an unlocked copy, and split that copy. Do that only when you are allowed to open the document in the first place; a password is sometimes there because the pages are not meant to travel.
Can I extract pages and send only those to someone?
That is exactly what this is for, and it is the strongest privacy move available: the pages you leave behind are not redacted, they are simply not in the file. Name the output so the recipient knows what it is, mention the source and range in your message, and offer the full document if they genuinely need it: sending only the relevant pages is normal practice with lawyers, landlords, courts and professors.