You open a contract, delete the page with the pricing annex, save, and email the file. The reader on the other end sees one page fewer. Whether the annex is actually gone depends on how the program saved the file, and it is entirely possible for the page to travel along, invisible in every reader, but present in the bytes and recoverable by anyone who knows where to look. The reason is a feature of the PDF format that exists for good reasons, and it is easy to work around once you know it is there.
What Deleting a Page Changes
ISO 32000, the PDF standard, stores a document as numbered objects. Each page is an object of its own, pointing at the content streams that hold its drawing instructions and at the fonts and images those instructions use. The pages are listed in a page tree, and the page tree is what a reader walks to decide what page 1, page 2 and page 3 are.
The quickest way to delete a page is to take its entry out of the page tree. Every reader will then show one page fewer. Nothing in that step requires the page object, its text or its images to be removed from the file. They are simply no longer listed. Whether they disappear depends entirely on what the program does when it writes the file back to disk.
Incremental Saves Keep the Old Version
Section 7.5.6 of the standard allows a file to be updated incrementally. Instead of rewriting the document, the program appends the objects that changed to the end of the file, followed by a new cross-reference section and a new trailer that points back to the previous one. The original bytes are left exactly as they were. The standard is explicit about deletions: an object that is deleted is left unchanged in the file and only marked as free in the new cross-reference section.
Incremental saving is not a flaw. It makes saving a small change to a large file fast, and it is what lets several people sign the same document one after another, because each signature covers the bytes as they stood and a later update must not alter them. But it means a file can hold several versions of itself. Each update ends with its own %%EOF marker, so cutting the file off at an earlier marker gives back the earlier version, deleted page and all.
Adobe Acrobat uses this approach for File > Save. File > Save As writes the document out again from scratch, folding the updates in and dropping what no longer belongs to it. The same edit can therefore leave the page in the file or remove it, depending on which of the two menu items was used.
Is the Deleted Page Still in the File?
How the way a file is saved decides whether a deleted page's content goes with it.
| How the file was saved | Deleted page content | Why |
|---|---|---|
| Incremental save (Acrobat's File > Save) | Still present | Old objects stay, only marked as free (7.5.6) |
| Page removed from the tree, whole file rewritten | Depends on the program | Unlisted objects are kept unless the writer discards them |
| Full rewrite that drops unused objects (Save As) | Gone | Only objects the document still uses are written |
| Kept pages copied into a new document | Gone | The deleted pages are never copied |
A rewrite is not enough on its own
Writing the file again in full does not guarantee that unlisted pages are dropped. The widely used open-source library pdf-lib writes every object it loaded when it saves. We checked this with a three-page test file: after removing the middle page and saving, the file had two pages, and the text of the removed page was still in it. The Delete PDF Pages tool here builds a new document from the kept pages for exactly that reason.
Delete pages into a fresh file
Mark the pages you do not want and download a new PDF built only from the ones you keep.
What a Fresh Copy Keeps, and What It Leaves Behind
Copying the kept pages into a new document is the reliable way to drop the deleted ones, and it changes a few other things along the way.
Kept: everything drawn on the pages
Text, images and vector graphics are copied, not redrawn, so text stays selectable and scans keep their quality. Page numbers printed in a footer stay what they were, because they are part of the drawing.
Kept, with a gap: links
Web links and links from one kept page to another still work. A link that pointed at a deleted page has nowhere to go and stops doing anything, rather than leading to a copy of the page that was supposed to be removed.
Dropped: document-level features
The bookmark outline, interactive form fields, attachments and digital signatures belong to the document rather than to any page, so a new document starts without them. If you need those, delete the pages in a full editor and use Save As.
Dropped: the edit history
Old versions from earlier incremental saves are not copied, so they cannot be recovered from the new file either. This is often why the new file is smaller than the page count alone would suggest.
Frequently Asked Questions
Related Tools
Keep Reading
Splitting a PDF: Why the Parts Weigh More Than the Whole
One shared font becomes twenty copies, a contents page can drag the whole document along, and edit history quietly drops out. Where the bytes go in a split.
Merging PDFs: What Comes Through and What Quietly Disappears
A merge copies pages, not documents. Text and images arrive untouched; bookmarks, form fields and signatures do not, and the reason is in the file format.
Rotating a PDF: Why the Page Turns Back When You Reopen It
The rotate button in most viewers turns your screen, not the file. How a page stores its orientation as one number, and why duplex scans flip every other page.