How to Repair a Corrupt PDF That Won't Open (Without Uploading It)
To repair a corrupt PDF, open the Repair PDF tool in your browser, drop the broken file, and download the rebuilt version. It runs qpdf — a widely trusted PDF engine — compiled to WebAssembly, entirely on your own device. A file that won’t open is often an important one, so nothing is uploaded, and it works with your Wi-Fi off.
When a reader says a PDF is “damaged” or “cannot be opened,” most people assume the document is gone. Usually it isn’t. The pages are still there — what’s broken is the map to them.
Fix it in your browser
- Open Repair PDF.
- Drop the file that won’t open onto the page.
- qpdf re-locates the surviving objects and rebuilds the file’s structure around them.
- Download the repaired PDF. Nothing was uploaded.
Why a PDF stops opening
A PDF keeps an internal index — a cross-reference table that says exactly where each object (page, font, image) sits in the file. Readers use it to jump straight to what they need. If those offsets are wrong, the reader gives up before it ever reaches the pages.
That index gets corrupted in ordinary, recoverable ways:
- Junk bytes bolted on. Mail gateways, download managers and “wrapper” scripts sometimes add bytes before or after the real document. Every internal offset is now off by that amount.
- A buggy generator. Some tools write a cross-reference table that doesn’t match where the objects actually landed.
- A rewrapped or re-saved file whose trailer no longer points at the current index.
In all of these, the pages survived intact. Repair re-finds the objects and writes a structurally clean file around them.
What repair can and can’t do
Repair is honest about its limits, and it helps to know them before you start.
It can fix: junk added before or after the document, object offsets that no longer match reality, and similar index-level damage. The content is present; only the pointers are wrong.
It can’t fix: content that is genuinely gone. If a download was cut off mid-transfer, the tail of the file — where the recovery anchor lives — may be missing, and there’s often nothing to rebuild from. Destroyed page or image bytes can’t be resurrected. Repair recovers what still exists; it doesn’t recreate data.
If repair can’t recover it
- Re-download or re-export the original. A truncated file is the common unrecoverable case — getting a complete copy is the real fix.
- Check your email’s “sent” copy or the source app. The version before it was mailed or wrapped is often fine.
- Try opening it in a different reader. Some are more forgiving of a bad index than others, which can buy you time to get a clean copy.
Related
- Compress PDF — shrink the repaired file if it’s now too big to email
- Organize PDF — reorder or drop pages once it opens again
- Unlock PDF — remove a password you know from a protected file
Frequently asked questions
What kind of PDF damage can this fix?
It rebuilds a broken cross-reference table — the index that tells a reader where each object is. That's the most common cause of a PDF that suddenly won't open; the actual content is usually intact.
What can't it fix?
If the file was truncated or the page data itself is genuinely gone, there's nothing to rebuild. Repair recovers structural damage, not missing content.
Is my broken file uploaded to repair it?
No. The rebuild runs in your browser with qpdf in WebAssembly, so even a sensitive damaged document never leaves your device.
Will I lose anything if the repair works?
Usually nothing — a successful rebuild produces a working copy with the same pages and text. Keep the original until you've confirmed the repaired file opens.