Extract pages from a PDF and keep only what you need

The quick way: pick the pages, save one PDF

1. Open Split PDF and choose your PDF. Every page appears as a card labelled Page 1, Page 2 and so on.
2. Keep "Pick pages", the first choice under "Split how".
3. Pick your pages. Tap or click the cards you want, or type them in the "Pages to take out" box, like 2-4, 9.
4. Leave "Save them as" on One PDF and press the yellow button. It says exactly what you will get, like "Save 4 pages as one PDF".
5. Download. The new file is named after your PDF and its pages, such as statement-pages-2-4_9.pdf.

That's all. Your original file isn't changed: "extracting" here means copying the pages you chose into a new PDF. The pages keep their full quality, because they are copied as they are, never redrawn. It is quick, too. In our tests, pages 41 to 45 of a 117-page textbook came out in 0.5 seconds as a 354 KB file, and pages 3 to 5 of a 14-page paper in 0.2 seconds.

Everything happens inside your browser. That matters for the papers people usually take pages from: bank statements, salary slips, ID scans and certificates. Not a page of them is sent to a server.

Find the right page numbers first

The most common mistake is taking the wrong pages, and it nearly always has the same cause. The splitter counts pages from the first page of the file, like the cards do. The numbers printed on the pages can be different: a book's cover and contents often have no number or use i, ii, iii, so printed page 1 might be page 9 of the file.

Always go by the label under each card. For a short file, just scroll through the cards. For a long one, find the page another way first:

• Search for a word. Open the file in our PDF Viewer and search for something on the page you want: "Class XII", "Annexure B", "April". The page box at the top shows where you are, like 37 / 120.
• Use the bookmarks. Many reports and books have a clickable contents list. The Viewer's Bookmarks tab jumps to each chapter, so you can note where it starts and ends.
• Note both ends. For "the April to June pages of my statement", find the first page that shows April and the last page that shows June. Then type the run, like 4-9.

Search needs real text. A scanned PDF is a picture of each page, so there is nothing to search; the Viewer tells you when a file has no text. For scans, scroll the cards and read the pictures.

Picking pages and ranges, the fast way

You can tap cards, type numbers, or mix both: the box and the cards follow each other. A picked card gets a yellow border and a tick. Tap it again to unpick it. On a computer, click the first page of a run, then hold Shift and click the last.

Typing is faster for anything longer than a few pages. A hyphen joins a run, commas separate the parts, and a run can be open at either end. Here is how common requests translate:

"All" picks every page and "None" clears them, which is handy when you want to start over. If the line under the box turns red, it says what's wrong, such as a page past the end of the file; the guide lists those messages in how to type page numbers.

One thing to know about order: Pick pages always keeps the pages in their original order. Typing 9, 2 gives page 2 then page 9. If you need a different order inside one file, extract the pages, then rearrange them with Organize PDF.

One file, or one file per page or range?

Before you press the button, decide what the person receiving the pages needs. Split PDF can give you any of these:

A single new PDF downloads as it is. Several come together in one ZIP file, a single file that holds others, named like report-split.zip. On Android, tap it in your Files app and choose Extract; on iPhone and iPad, tap it in the Files app to unpack it. The guide has the file names and ZIP steps for computers too.

If a portal wants each document as a separate upload, A PDF per page is the quickest route, but rename each file before uploading. A name like statement-page-3.pdf says nothing about what's on it; PAN-card.pdf does.

What happens to bookmarks, links and forms

A PDF holds more than what you see on its pages. When you take some pages out, the rule is simple: whatever points at a page you kept stays and works; whatever points at a page you left out is removed, since it would lead nowhere.

• Bookmarks (the clickable contents in your PDF reader's side panel) stay when they point at kept pages. When one is removed, the bookmarks under it move up to take its place.
• Links to kept pages still work, and so do links to websites. Links to pages you left out are removed.
• Form fields on kept pages stay fillable. Fields on pages you left out go with them.
• Tags and page labels are dropped when pages are left out. Tags are the hidden structure screen readers follow; page labels are custom numbers like i, ii, iii.

We measured this on our 117-page maths textbook, which has 35 bookmarks and 509 links. Cut into two halves, each half kept the bookmarks for its own pages: 17 in one and 18 in the other. 316 of the links, those leading within the same half or to websites, still worked. The 193 links that led into the other half were removed. In another test, splitting a two-page form into single pages kept each page's 8 fields, still fillable.

In practice: if you take out a contents page on its own, its text stays but its entries stop being clickable, because their chapters aren't in the new file. Take the chapters with it and the links keep working.

A privacy note. Fonts and pictures used only by pages you left out are not copied, so the other pages' content doesn't hide inside your extract. The document's details, such as its title and author, do go into every new PDF, along with files attached to the document. If the title says something private, such as a client's name, check it in your PDF reader's document properties before you share the extract.

How big the extracted file will be

A few pages from a long PDF make a small file, because only what those pages need is copied: their text, fonts and pictures. From our tests in Chromium 141, the browser Chrome is built on:

Big files need memory as well as time: the browser used about 1.1 GB more memory than when idle for the 400-page scan, and about 2.7 GB for the 1,000-page one. Use a computer for files that size.

Scanned pages are pictures and take far more room than typed ones. If your extract is still too big for a form, make it smaller with our PDF compressor, or with Resize PDF when the form gives an exact limit, like under 200 KB.

Extracting pages on a phone

It works the same in Chrome on Android and Safari on iPhone and iPad, with no app to install. Tap "Choose a PDF file" and pick it from Downloads or the Files app. A PDF that arrived on WhatsApp or by email may need saving to your files first.

• Type runs instead of tapping. A touch screen has no Shift-click, so for pages 12 to 30, typing 12-30 beats thirty taps.
• Let the pictures load. Each card's picture is drawn as it comes into view, so a long file fills in as you scroll. You can pick pages by number before the pictures appear.
• The button stays in reach. It sits at the bottom of the screen, so you don't have to scroll back up to press it.
• Find the file afterwards in Downloads, in your Files app. On iPhone, that is where Safari usually saves it.

Phones have less memory than computers. Everyday statements and certificates should be no problem, but for scans of hundreds of megabytes, use a computer.

When deleting pages is the better route

There are two ways to end up with "only some pages": pick the ones you want, or delete the ones you don't. Split PDF picks; Organize PDF shows every page as a card with a delete button. Choose by the job:

Deleting also makes a file smaller when the deleted pages are scans. In our test, deleting the last 10 pages of a 42-page colour scan (and swapping two others) took it from 47.8 MB to 36.4 MB in 0.6 seconds, because the deleted pages' pictures were left out.

Other ways to do it, and their catches

• Print to PDF with a page range. In most print windows you can choose which pages to print, then pick Save as PDF (in Chrome) or Microsoft Print to PDF (on Windows) as the printer. It's fine for a quick page or two. But printing makes a new document from what is drawn on the page, so bookmarks and fillable fields usually don't come along.
• Preview on a Mac. You can drag page thumbnails out of Preview's sidebar to make a new PDF of those pages. It's built in and works offline.
• Online PDF websites. They work, but most of them upload your file to their servers to do it. For a bank statement or an ID card, that's a copy of your document on someone else's computer.

Split PDF copies the pages themselves, so text stays text, links and bookmarks to kept pages keep working, and form fields stay fillable, without the file leaving your device.

Problems and quick fixes

• The wrong pages came out. You probably used the numbers printed on the pages. Check the labels under the cards and try again; your original is unchanged.
• The pages aren't in the order I typed. Pick pages keeps the original order. Use By range for separate files in your order, or Organize PDF to reorder one file.
• "Locked with a password." The file needs a password to open. Remove the password in your PDF reader, then add the unlocked copy. A file that opens without a password but blocks editing is split as normal.
• The extract of a signed document shows no valid signature. A signature only covers the file as it was signed, and an extract is a changed copy. Send the original as well if someone must check it.
• The new file is too big for a form. Compress it, or resize it to the exact limit.
• You need pages from two PDFs in one file. Extract from each, then join them with Merge PDF.

Cutting a big file into many parts rather than taking a few pages out? Splitting a large PDF into smaller files covers every N pages, upload limits and huge scans.