It is a blunt message with an exact meaning. The drive read a sector, compared it with the checksum stored beside it, and the two did not agree. That is a reliability problem in the hardware rather than a file you can repair your way out of, and the tools Windows offers next are the ones most likely to make it permanent.
A CRC error is a reliability fault, not a file fault. Free diagnostic, one written figure, £300 + VAT for a single drive or SSD — and no software step until the cause is known.
Data error cyclic redundancy check is a blunt message with a precise meaning, and it turns up in front of several unrelated problems. What separates them is where the error appears and what else the drive is doing.
Drag a folder across and the transfer halts at the same point with the same message. That points at one specific patch of unreadable sectors rather than a drive in general trouble, and it is usually the earliest warning anybody gets.
The volume opens, folders list normally, and nothing will actually copy. When failures are scattered across the whole surface rather than gathered in one place, the disk is degrading generally and the only priority is getting an image off it.
A CRC error accompanied by ticking, a faint scrape or a change in pitch is not a media fault you can read your way past. Damaged heads produce unreliable reads and go on damaging the surface for as long as the drive has power.
A failing USB lead, a tired bridge board in a portable case or a marginal SATA connection corrupts data in transit and the checksum catches it at the far end. Connect the same disk another way and the error disappears. Cheap to rule out, so rule it out first.
Losing power partway through a write leaves sectors half finished, holding contents that no longer match their stored checksum. The mechanism is usually unharmed and the file system is the thing that needs rebuilding, which is logical work.
Windows offers to scan and repair, and on a physically failing disk that offer is expensive. The scan reads weak areas again and again and writes its corrections onto a drive that has already proved unreliable. This is how recoverable jobs become difficult ones.
A cyclic redundancy check is a checksum, nothing more mysterious than that. When a sector is written, the drive calculates a short value from the contents and stores it alongside them. When that sector is read back, the value is calculated again and compared with the one held on the disk. Matching values mean the data survived intact. Differing values mean something has changed since the write: the magnetic signal on the platter has weakened, the head is not reading it cleanly, or the bits were corrupted somewhere between the disk and the computer. Windows reports the mismatch as data error cyclic redundancy check and stops, because handing you data it knows to be wrong would be worse than refusing to hand you anything. The message is a symptom of unreliable reads. The only useful question is what is causing them.
The first is surface degradation. Sectors that once held a strong signal weaken, or the head reading them is worn or slightly damaged, and the drive makes repeated attempts before it either succeeds or gives up. You hear this as a machine that hangs for half a minute at a time and see it as copies that stall at the same place. This is the serious version and it gets worse with every hour of use. The second is a transmission fault. A frayed USB lead, a loose SATA connector, a bridge board in an external case losing its grip: the disk reads the sector correctly and the data is mangled on the journey. Swap the cable, move the drive out of a hub and into a port on the machine itself, and try it on a different computer if there is one about. If the error travels with the cable you have found it. If it stays with the drive, stop testing and start planning.
Windows will suggest scanning and repairing the drive, and on a healthy disk with a muddled file system that is fair advice. On a disk with failing sectors it does active harm, for two separate reasons. CHKDSK reads the entire surface, which means it returns to the weakest areas repeatedly and asks a struggling head to try again, and again, and again. Heads and platters both suffer for that, and drives have died halfway through a scan. It also writes. Marking clusters bad, rebuilding directory entries, dropping what it thinks it has rescued into numbered found folders: all of it lands on a disk that has already demonstrated it cannot be trusted to store things properly. Files come back truncated, renamed to digits, or not at all. If a scan is running right now on a drive you care about, the correct move is to power the machine off rather than let it finish.
One rule covers every version of this error. If the drive is making any noise it did not make a month ago, no software step applies to it. Take the power off and send it. Software cannot mend a mechanism, and running it costs you the reads a bench would otherwise have used.
Nothing on this bench argues with a failing sector. The drive is attached to a hardware imager that controls it directly rather than through Windows, so read timeouts are set in milliseconds instead of the many seconds a normal controller spends retrying. The imager takes the healthy regions first and at speed, building a map of where it met resistance. It then returns for the difficult areas in separate passes, using smaller block sizes, reads run in reverse, head-by-head selection on drives with several, and rest periods so the mechanism cools between attempts. What ends up on our storage is a sector-accurate copy of everything the disk was willing to give up, and all file system repair happens on that copy. Your drive is never written to at any stage. Where a disk is too weak to survive imaging as it stands, the mechanical work at the clean bench comes first.
A CRC error can turn out to be either sort of job, which is exactly why the diagnostic comes first and costs nothing. A logical case, such as sectors corrupted by a power cut or a file system needing repair on healthy hardware, is £300 + VAT for one drive or SSD and sits under no fix, no fee. A drive with genuine surface or head damage is mechanical, and the published rule is precise: no fix, no fee all jobs except electronic and mechanical failures, chip level work, DVR and Forensic jobs. Those take 50% of the quoted price upfront, agreed in writing beforehand. There is no collection service anywhere in this network. Post the drive to Manchester Data Recovery, Peter House, Oxford Street, Manchester M1 5AN, which is around 35 miles from Liverpool and next working day by tracked post, and the free diagnostic closes 2 working days after it arrives. If the same drive is also ticking, read the clicking hard drive page first, and if Windows now calls the volume RAW, the RAW drive recovery page is the closer match.
Stop the scan, take the power off, and let the imager read around the damage instead of through it.
Nearly everything on this bench arrived by tracked, insured post. It is the calmest way to move a failing drive, and a parcel handed in anywhere on Merseyside is normally on the bench the next working day.
Still screwed inside a laptop, tower, MacBook, iMac, server or a CCTV recorder? Take the drive out and post that on its own. We do not strip machines here, and any computer shop will do it in a few minutes. The one job nobody can take on is flash soldered straight to a logic board, as on Apple Silicon Macs and a handful of very slim laptops: if the storage will not unbolt, there is nothing to send.
↓ Print the shipping & booking-in form (PDF)
Address it to Manchester Data Recovery. It is roughly 35 miles from Liverpool along the M62 if you fancy the run, and next working day by tracked post if you do not. We ring or email the moment it is booked in, and the free diagnostic closes 2 working days after that.
Not sure what belongs in the parcel? Call 0800 689 0668 before you tape it shut, or step through the free online diagnostic first.