You open a folder or double-click a file and Windows says: “The file or directory is corrupted and unreadable.” Sometimes it arrives as error 0x80070570 instead. Sometimes the whole drive throws it; sometimes just one folder.
The wording sounds final. It usually isn’t. This error is about the index, not the contents.
What’s actually broken
NTFS keeps a Master File Table — a record for every file and folder on the volume, holding the name, the timestamps, the permissions, and the map of which sectors hold the actual data.
Your files live in those sectors. The MFT just tells Windows where to find them.
When Windows says corrupted and unreadable, it means it read an MFT record or a directory index entry and got something structurally invalid. It can’t follow the map. The territory — the actual bytes of your photos and documents — is almost always sitting there untouched.
That distinction is the whole reason these cases usually end well.
How the index gets damaged
Unclean shutdown mid-write. Power loss or an unsafe eject while metadata was being updated leaves a half-written record. The most common cause on external drives by a mile.
Bad sectors landing on metadata. The MFT occupies a small fraction of the disk, but it’s the most important fraction. A handful of failing sectors in the wrong place take out records for thousands of files while 99.9% of the drive reads perfectly. If this is the cause, you’re also dealing with a physically degrading drive and the priority changes.
Interrupted resizing or partition work. Partition tools that got interrupted, or a cloning job that stopped halfway.
Failing USB bridge on an external. A flaky controller can corrupt writes in transit, producing metadata damage on a drive that’s mechanically perfect. Worth ruling out — enclosure failures look a lot like drive failures.
The fix everyone recommends is the wrong one
Search this error and you’ll be told to run chkdsk /f. Windows itself will offer to “Scan and fix.”
Understand what that does here. The damage is inconsistent metadata. CHKDSK’s method for resolving inconsistent metadata is to discard the parts it can’t reconcile. Files whose records don’t validate get stripped of their names and their place in the folder tree and dumped into found.000 as numbered .chk files.
Your data survives. Everything that made it findable does not. Recovering names and structure afterward is possible but much harder than it would have been from the damaged original — CHKDSK overwrote the very records a recovery tool would have used to rebuild the tree. The full explanation is here.
Free evaluation. Firm quote before any work begins.
What proper recovery looks like
Image the drive sector by sector first, then work on the copy.
From the image, the MFT gets parsed directly rather than trusted. NTFS helpfully keeps a partial mirror of critical MFT records, and file records are highly structured and recognizable even in isolation — so damaged entries can often be reconstructed from the mirror, from the journal, or from surrounding records. When a folder’s index is gone but its files’ records survive, the folder tree can be rebuilt around them.
These recoveries frequently come back complete: filenames, folder structure, timestamps. It’s the same category as a RAW volume — a navigation problem, not a data problem — and it’s the most favorable kind of case I see.
The bottom line
“Corrupted and unreadable” describes the file system, not your files. The files are nearly always intact behind a broken index. Don’t let anything repair that index in place — image it, rebuild from the copy, and you’ll usually get everything back with its name still attached.
Seeing this on a drive that matters? Start a case — and tell me whether it hits one folder or the whole volume. A single folder points at localized damage; the whole volume points at the MFT itself.