Optimize PDF · guide
What Repair PDF does
Repair PDF rebuilds a PDF that will not open, or opens with errors, into a new working file. It copies every page it can read into a fresh, correctly indexed document, which fixes the common damage: a broken cross-reference table or a file an editor saved badly, where the pages are intact and only the index that finds them is broken. A page Repair finds but cannot copy is dropped and named in the result. A page whose data is missing altogether is left out without being named, so compare the page count with the original. The pages come across, but bookmarks, document attachments, the title and author, the document language, page labels, accessibility tags and layer settings do not survive the rebuild, so anything on a hidden layer becomes visible. The file never leaves your browser.
How to repair PDF in your browser
- Choose the damaged PDF Select the PDF that will not open properly. It is read in this browser tab and not uploaded.
- Repair it Select Repair. Every page that can be found is copied, one at a time, into a fresh, correctly indexed file.
- Check and save Read the page count and any pages the result names as lost, compare the count with the original, open the repaired copy and look through every page, then save it.
Worked example
Recover readable pages from a PDF with damaged structure.
- Starting point
- Choose a file that a viewer reports as damaged, and preserve an untouched copy.
- What to do
- Run Repair, inspect reported page count and notes, then reopen the result in another PDF viewer.
- Result
- Expect a rebuilt PDF when readable page contents can be recovered.
If the result is not right
If pages or content are missing, stop using the result as a complete record. Obtain another source copy or seek recovery while retaining the damaged original.
Limits worth knowing
- Repair cannot recreate missing bytes. Repair counts only the pages it can find, so compare the page count with the original. Bookmarks, document attachments, the title and author, the document language, page labels, accessibility tags and layer settings do not survive this page-recovery workflow, so anything on a hidden layer becomes visible.
- This workflow stops on interactive forms rather than risk losing their fields. Use Flatten first and check every visible answer in that saved copy, then open the flattened PDF here. Keep the editable original; protected or XFA forms may need a compatible form application first.
Questions and answers
Can every broken PDF be repaired?
No. Repair recovers files whose pages are intact but whose index is damaged, which covers most files an application saved badly. A download cut short can only give back the pages that arrived, and none at all if the part that lists the pages is missing. If Repair finds a page but cannot copy it, that page is dropped and the result lists it. A page whose data is missing altogether is left out without being named, so compare the page count with the original, and ask the sender for a fresh copy if pages are short.
My PDF says it is damaged after downloading. What should I try first?
Download it again, since a download cut short is the most common cause. If the file is only a few kilobytes, it may not be a PDF at all: some sites send an error page under a .pdf name, and there is nothing inside to recover. If a fresh download still fails, run it through Repair PDF.
Why is a page blank or missing in the repaired copy?
A page Repair finds but cannot copy is dropped so the rest of the file opens, and the result lists it by number. A page whose data is missing altogether is left out without being named, so the page count is the only sign of it: compare the count with the original. A page that comes through but looks blank or scrambled was found and copied, but the drawing instructions inside it were damaged or cut short, and Repair cannot redraw them; the last page of a cut-short file can be one. That is why it is worth looking through every page before you rely on the copy.
Can it repair a password-protected PDF?
Not while it is locked. Repair opens the file without a password, so a PDF that needs one fails, and the message can read like damage. Open it in Unlock PDF with the password first, then repair the unlocked copy; if Unlock cannot open it either, ask the sender for a fresh copy.