Cyclic redundancy check error deserves a clear, practical explanation. Of all the errors Windows throws at you, “Data error (cyclic redundancy check)” is one of the more honest. It is not a vague complaint. It means the drive read a block, compared it against the checksum stored alongside it, and the two did not match. Something between the platter and the cable changed the data.
The short answer. CRC errors mean a physical or connection problem, not a software one. Try a different cable and port first, because a bad SATA or USB cable produces identical symptoms to a failing drive. If it persists on good cables, the media is degrading and your priority is copying files off, not repairing anything.
Cyclic redundancy check error: Cable first, always
It costs two minutes and it resolves a meaningful share of these. A marginal SATA cable, a USB cable that has been coiled tightly for two years, or a drive on an unpowered hub will all corrupt data in transit and report CRC errors from a perfectly healthy drive. Swap the cable, use a rear port on a desktop rather than a front one, skip the hub entirely.
Then check SMART
- Reallocated Sector Count above zero and rising — the drive is remapping bad sectors. Replace it.
- Current Pending Sector Count above zero — sectors it cannot read and has not yet remapped. These are exactly your CRC errors.
- UDMA CRC Error Count rising — this one is the cable, not the platter. Genuinely good news.
- Uncorrectable Sector Count above zero — data in those sectors is already lost. Copy the rest now.
That UDMA CRC distinction is worth knowing. It is the only SMART attribute that points away from the drive, and a lot of people replace a healthy disk because nobody told them.
Do not run chkdsk /r on a failing drive
This is the standard advice everywhere and it is wrong when the drive is dying. chkdsk /r reads and rewrites every sector on the volume, which on a drive with degrading media means hours of hammering the exact hardware that is already failing. Drives have died mid-chkdsk and taken everything with them. If SMART is clean and you have a backup, fine. Otherwise, copy first.
Copy in the right order
| Priority | What to copy | Why |
|---|---|---|
| 1 | Irreplaceable personal files | Photos, documents, anything with no other copy |
| 2 | Work in progress | Recreatable but expensive |
| 3 | Everything else, smallest first | More files rescued before the drive gives up |
| Never | The whole drive as one big copy | One bad sector aborts the entire operation |
If a normal file copy stalls or aborts when it hits a bad sector, that is the point to switch to recovery software. A copy tool gives up on an unreadable block; recovery software notes it, skips it, and carries on with the rest of the file. You lose the bad block and keep everything around it.
Copying past the bad sectors
EaseUS Data Recovery Wizard can create a sector-by-sector image of a drive that is throwing CRC errors, so you get one careful pass at the failing hardware and do the rest of the work on the healthy copy. That is the right approach when the drive may not survive a second attempt.
When it is an optical disc or a downloaded file
CRC errors on a DVD or on an extracted archive mean something different: the disc is scratched or the download was incomplete. For archives, re-download — do not try to repair. For discs, a different drive often reads what yours cannot, because optical drives vary enormously in error correction.