Data Recovery

"Data Error (Cyclic Redundancy Check)": What It Means for Your Files

You’re copying files off an external drive and it stops partway: “Data error (cyclic redundancy check).” Maybe it happens on one file every time. Maybe it happens somewhere new on every attempt.

This error is more specific than most Windows messages, and what it’s telling you is worth understanding before you decide what to do next.

What a CRC error actually is

Every sector written to a drive carries a checksum alongside the data. On read, the drive recomputes that checksum and compares. Match means the data came back intact; mismatch means it didn’t.

A cyclic redundancy check error is the drive saying: I read this sector, and what came back doesn’t match what was supposed to be there. The drive tried, retried internally, and gave up.

That’s a hardware-level report, not a software complaint. It’s the difference between “Windows is confused about this file” and “the physical media could not return this data correctly.”

What causes it

Degrading sectors. The most common cause by far. The magnetic layer weakens over time, or the head can no longer read a particular region reliably. This is normal wear that becomes visible once it outpaces the drive’s spare-sector pool.

A failing head. If errors cluster on certain regions rather than scattering randomly, one head in a multi-head drive may be weakening. That’s a different, more serious diagnosis.

Cable, port, or enclosure. Less common, but real — a marginal USB bridge or a bad cable can corrupt data in transit and produce the same error. This is the one benign possibility, and it costs nothing to rule out: try a different cable and a different port directly on the machine, not through a hub.

Interrupted writes. A file that was being written during a power loss can carry a bad checksum permanently, even on a healthy drive.

Why “retry” and CHKDSK make it worse

The natural response is to run the copy again. And again. Each attempt sends the head back to the same weak region for another round of retries, and the drive doesn’t fail politely — it hammers.

The standard internet advice is chkdsk /f /r, which is the escalated version of the same mistake. The /r flag forces a read attempt on every sector on the disk, retrying each weak one. On a drive that’s already returning CRC errors, that’s hours of maximum stress on failing hardware, and I regularly see drives that were readable before the scan and weren’t after. More on why CHKDSK is the wrong tool here.

Start a Recovery Case

Free evaluation. Firm quote before any work begins.

What to do instead

If the files matter, stop copying and power the drive down.

The right approach is a sector-level image made by software designed for failing media — it reads in large blocks first, skips over problem areas instead of grinding on them, and comes back for the hard sectors afterward with the healthy 99% already safely captured. That ordering is the whole trick. Consumer copy tools do the opposite: they hit the first bad sector and stall there, stressing the drive while everything past it stays uncopied.

Once the image exists, the bad sectors are a known, bounded problem. A handful of unreadable sectors usually means a few damaged files out of thousands, and the rest come back clean.

Is a CRC error a death sentence?

No — but it is a warning with a clock on it. A drive throwing CRC errors is a drive whose spare-sector pool is exhausted or whose media is degrading, and that trend doesn’t reverse. Treat it as a drive to get data off of, not a drive to keep using.

If you got the error while copying and the copy otherwise mostly worked: finish getting your data somewhere safe, then retire the drive. Don’t reformat it and put it back in service. A format doesn’t repair anything physical — it just hides the evidence until next time.

The bottom line

CRC means the media couldn’t return the data correctly. It’s the drive being honest with you. Repeated retries and CHKDSK both respond by stressing the exact hardware that’s failing. Image it once, carefully, and work from the copy.


Getting CRC errors mid-copy? Start a case — mention whether the error lands on the same file every time or moves around. Consistent means one bad region; moving means something broader, and it changes the plan.

Dealing with this right now?

Describe what's happening and I'll let you know if I can help. No charge to find out.

Start a Recovery Case