Why Won't My PDF Open? What "Repairing" a PDF Actually Does
Published 9 August 2026 · 5 min read
A file that opened fine yesterday suddenly throws "Failed to load PDF document" today, or it arrived that way after a bad download or a USB transfer that got interrupted halfway. The panic is usually proportional to what's in it — a signed contract, a scanned tax document, the only copy of something you needed an hour ago. Before assuming it's gone, it's worth understanding what actually breaks inside a PDF, because "corrupted" covers a wide range of problems, and a lot of them are genuinely fixable.
What actually goes wrong inside a "corrupted" PDF
A PDF file is really a collection of objects — pages, fonts, images, text streams — plus a cross-reference table that acts as an index, telling a PDF reader the exact byte offset where each object lives. Most real-world corruption comes from something interrupting a save or transfer partway through: a browser crash mid-download, a connection that dropped during an upload, a drive that disconnected during a copy. What usually breaks first is that index, or the very end of the file where it lives — which is why a PDF reader gives up immediately rather than showing a partially broken page. The content is very often still sitting intact in the file; the reader just doesn't know where to find it anymore.
Why "just open it in another app" sometimes works, sometimes doesn't
Some PDF readers are more forgiving than others about a damaged cross-reference table — a few will attempt to scan the raw file for recognizable objects even when the index is broken, which is why a file that fails in one app sometimes opens in another. That's not a repair, though — it's a reader tolerating the damage on that specific attempt, not fixing the file for next time or for whoever you send it to.
What an actual repair tool does
Genuine PDF repair works the way the more forgiving readers do internally, but saves the result: it scans through the raw bytes of the file looking for valid PDF objects — pages, fonts, image data, text streams — even without a working index to point to them, then rebuilds a clean, valid cross-reference table and writes out a new file that any standard PDF reader can open normally. It's reconstruction from the pieces that are still there, not a trick that only works in one particular app.
How FormatDog's Repair PDF avoids uploading it in the process
FormatDog's Repair PDF tool performs that same raw-object scan and cross-reference rebuild using a PDF library running as JavaScript in your browser — the potentially-corrupted file is read locally, reconstructed locally, and the repaired result is handed back to you without ever being uploaded anywhere in the process. For a document you're already worried might be damaged, that also means no wait on an upload of a large scanned file, just to find out afterward whether the repair worked.
You don't have to take that on faith
Open your browser's developer tools (F12, or right-click and choose "Inspect"), click the Network tab, then run the repair tool on a file. If it's uploading the PDF to attempt the repair server-side, you'll see an outgoing request carrying it appear. If the repaired file downloads with nothing showing up there, the whole reconstruction happened locally.
The honest caveat
Not every corrupted PDF is recoverable — if the actual object data (not just the index) was overwritten or the file was truncated mid-write, there may genuinely be nothing left to reconstruct. Repair tools, local or server-side, can only rebuild from what's still physically present in the file. For most "won't open" cases caused by an interrupted transfer, though, the content is intact and recoverable; it's the map that broke, not the territory.