PDF Won't Open? How to Repair a Corrupted File

You double-click a PDF and instead of the document you get an error: the file is damaged, or cannot be opened, or simply nothing happens. It is a bad moment, particularly when it is the only copy of something you need. The good news is that a surprising share of these files are repairable, because the damage is usually to the file's index rather than to its contents.

What is actually broken

A PDF is not one continuous stream of text. It is a collection of separate objects — pages, fonts, images, metadata — plus a table at the end of the file that says where each object lives. Your reader uses that table to find things.

When a PDF “goes corrupt”, that table is usually what broke. The pages are still sitting in the file, intact and complete, but the map to them is wrong, so the reader cannot locate anything and refuses to open the document at all. That is why repair works as often as it does: the content did not go anywhere, and rebuilding the map by scanning the file for objects restores access to it.

How files get damaged in the first place

• An interrupted download. The most common cause by far. The connection dropped, and you have most of a file with no valid table at the end — the table is written last.
• An interrupted transfer or save. A USB drive pulled out mid-write, a network share that dropped, a laptop that shut down while saving.
• Email truncation. Some mail systems silently cut attachments at a size limit rather than rejecting them.
• Failing storage. An ageing drive or a worn memory card writing bad sectors. If this is the cause, other files are usually affected too — worth checking.
• A faulty export. Software that crashed part-way through writing the PDF, or an unusual generator that produced a technically invalid file that some readers tolerate and others reject.

Before you repair: check it is really the file

A few minutes here saves effort, because not every failure to open is corruption.

1. Try a different reader. Open it in a web browser as well as your usual PDF application. A file that opens in one and not the other is often valid but unusual, not damaged.
2. Check the file size. Zero bytes, or far smaller than expected, means the content is not there. Repair cannot help; find another copy.
3. Download it again if it came from the web or email. An interrupted download is the single most common cause, and a clean re-download fixes it outright.
4. Check it is actually a PDF. Files occasionally arrive with the wrong extension — an image or a Word document renamed to .pdf will never open as one.

Repairing the file

If it is genuinely damaged, Repair PDF rebuilds the internal structure: it scans the file for whatever objects are still intact, works out the page order, and writes a fresh, valid document around them.

Add the damaged file, run it, and download the result. Then open it and check properly — do not assume a downloaded file is a fixed file. Confirm the page count matches what you expect, and look at the pages that matter rather than only the first one.

What a repaired file looks like

Outcomes fall into three rough groups, and it is worth knowing which you are in.

• Fully recovered. The index was broken, the objects were fine, everything is back. This is the most common result for interrupted downloads.
• Partially recovered. Most pages return, some are missing or blank. This usually means the file was truncated, so the later pages genuinely were not there to recover.
• Structure only. The document opens but pages render oddly, fonts are substituted, or images are missing. The page objects survived but things they referenced did not.

Partial recovery is still worth having. Nine pages of a ten-page contract is a far better position than nothing, and it tells you precisely what you need to source again.

When text comes back as gibberish

Sometimes a repaired PDF displays correctly but its text is unusable — copying produces nonsense, and searching finds nothing. This happens when the font data that maps shapes back to characters was among the casualties. The page draws correctly because the shapes survived; the file's record of what those shapes mean did not.

There is a way around it. Because the page still renders correctly, OCR can read it visually and write a fresh, correct text layer — exactly as it would for a scan. You are recovering the text from the picture of the page rather than from the file's broken internals.

If only part of the document survived

When some pages are damaged and others are clean, salvage what is good before doing anything else. Extract Pages pulls out the pages that survived into their own file, so the intact content is safely separated from the broken material.

If you later obtain a replacement for the missing section, Merge puts the document back together in the right order.

When it genuinely cannot be saved

Repair rebuilds structure around content that still exists. It cannot recreate content that was never written to the file. If the download stopped at ten percent, ninety percent of the document is not damaged — it is absent, and nothing will bring it back.

Signs you are in that situation: a file drastically smaller than it should be, a repair that returns only the first page or two of a long document, or repeated repairs producing the same partial result. At that point the effort is better spent finding another copy — the sender's outbox, a mail server, a backup, or the system that generated it originally.

Reducing the chance of it happening again

• Verify large downloads by opening them once, immediately, while you can still easily fetch the file again.
• Eject drives properly rather than pulling them out. Write operations are often still finishing after the copy dialog closes.
• Do not work on files directly from a USB stick. Copy to the computer, work there, copy back.
• Send large documents as a link rather than an attachment where you can, which avoids mail systems truncating them.
• Keep the original of anything important. A repaired file is a reconstruction, and reconstructions can carry subtle problems that only surface later.

A note on the repaired copy

Once you have a working file, treat it as the new master and store it somewhere safe before doing anything else with it. It is easy to repair a document, immediately start editing it, and end up with no clean copy of the recovered version at all.

Nothing you upload here is kept once your download is ready, so the repaired file exists only where you save it. That is deliberate — damaged files are often the important ones, and the last thing they need is another copy sitting on someone else's server.

If the file matters and repair only partly worked

Before giving up on a partially recovered document, check whether the content exists somewhere other than the file itself. Documents leave more copies behind than people remember.

• The sender's sent folder usually still holds the original attachment, intact.
• Mail servers often keep a copy independent of your local mailbox, reachable through webmail even when the downloaded attachment is broken.
• The system that generated it. Invoices, statements and payslips can nearly always be re-issued from the portal or software that produced them, and that copy will be perfect.
• Cloud storage version history. If the file lived in a synced folder, an earlier undamaged version may still be there.
• Another recipient. If the document went to several people, someone else's copy may have arrived intact.

A perfect original from any of these beats a partial repair every time. Repair is the right answer when no other copy exists — it is rarely the fastest answer when one does.

FAQ

Can every corrupted PDF be repaired?

No. Repair rebuilds the structure around content that is still present in the file. If the document was truncated during download or transfer, the missing part is absent rather than damaged and cannot be recovered.

Why does my PDF say it is damaged?

Usually because the internal table that tells the reader where each page lives is broken — most often from an interrupted download or save. The pages themselves are frequently intact, which is why repair often works.

Will I lose content when repairing?

Often nothing at all. If the file was truncated you may get a partial document, with the pages that survived recovered and the rest missing.

The repaired file opens but the text is nonsense. What now?

The font mapping was damaged. Run OCR on the repaired file — it reads the page visually and writes a fresh text layer, bypassing the broken mapping entirely.

Is it safe to upload a damaged file?

It is processed and discarded the moment your download is ready — never stored, never queued, never read by a person.

Why does my PDF open in a browser but not in my reader?

That usually means the file is valid but unusual rather than corrupt. Different readers tolerate different deviations. Repairing it produces a cleaner file that more readers accept.

Is repairing free?

You get five free uses a day with no payment. Pro removes the daily limit.