Inside the lab follow MDrepairs on YouTube, TikTok & Instagram - visit our YouTube channel ↗

CHKDSK Stuck or Made Things Worse? Stop and Read

Our founder and lead technician, Joseph Montanti, handles cases personally.

If CHKDSK gets stuck at one percentage, check whether its file count is advancing and whether the drive is behaving normally. A full scan of a large hard drive can take hours, including long periods at one percentage. Read errors, a drive disappearing from Windows or clicking from a hard drive change the decision: preserve the data before more repair attempts.

On this page

CHKDSK at 0 percent or stuck on stage 4

CHKDSK checks a volume's file-system records. The /f option repairs errors; /r includes those repairs and attempts to recover readable information from bad sectors. That extra reading is why /r can take much longer than a check of the directory structure. Microsoft documents these differences in its CHKDSK reference.

At stage 4, “Looking for bad clusters in user file data” describes the file-data check. On a large volume, a changing processed-file count can sit beside “0 percent completed.” A scan can also pause at 10% while the file count advances. There is no universal three-hour deadline after which a scan becomes a failure.

A read that times out may have to be retried, slowing the operation. Repeated failures can make CHKDSK appear frozen.

If the hard drive clicks, beeps or grinds, shut it down now and leave it off. Repeated spin-downs, a drop or liquid exposure are also reasons to leave it off. If SMART reports a failing disk, skip software scans and repairs.

What you seeWhat it means for the next step
Large drive, backed-up data, changing counts and no errorsAllow the scan time to finish.
Unchanged percentage but increasing file countThe displayed percentage is not the only progress indicator.
Read failures or repeated disconnectsResolve the storage fault and preserve the files before further repair.
Important files exist only on this driveCancel the running repair if it responds, then copy the files or image the drive.

Should you cancel the running scan?

Microsoft discourages interrupting CHKDSK, while explaining that interruption should not leave the volume more corrupt than before the run. That is more limited than saying cancellation is always harmless. If the scan is making progress on stable hardware and your files are backed up, letting it complete is reasonable.

Stop a repair that's running on your only copy. Cancel if the scan responds. If the scan freezes and the computer stops responding, shut it down.

Once the scan stops, leave the source unchanged. Don't install recovery tools there, save files back to it or restart CHKDSK with more switches. If Windows can still open the folders normally, copy the needed files to a different drive before further troubleshooting. End that transfer if reads begin failing or the device drops out. Seagate's Backup Plus Hub manual also puts backup before Check Disk.

When files went missing after CHKDSK

File-system repair changes the records that describe files and folders. On a damaged volume, making those records consistent can leave files without their former names or directory entries. The repair may shorten an already damaged file or mark bad clusters so Windows won't allocate them again. Microsoft specifically warns that FAT repairs can cause data loss and describes saving lost chains as File<nnnn>.chk files.

If you now see a FOUND.000 folder or .CHK files, preserve them along with the rest of the volume. They may contain file contents recovered from damaged records. Renaming every fragment to .jpg will not reconstruct missing pieces or restore the original folder structure.

On a drive that reads steadily, GNU ddrescue can copy the unmounted source into an image file on another disk. Use a mapfile and check the destination before starting. If reads time out, stop home imaging. Recovery can then work from that copy without repeatedly changing the source's file-system records. Our hard drive recovery work includes imaging; readable regions can be copied while damaged areas are handled separately. On a drive with hardware trouble, getting that image can itself require repair.

If CHKDSK finished and the volume now appears RAW or Windows will not boot, keep the current state for recovery.

An SSD can stall without making a sound

An SSD has no spinning platters or moving heads. CHKDSK still reads its logical storage and, with repair options, changes file-system records. Marking a cluster bad is a file-system action; it does not replace the flash controller's physical memory management.

A disappearing SSD, repeated read errors or timeouts need investigation even though the drive is silent. In Event Viewer, Windows Logs, System, look at errors from the time of the scan. Microsoft describes event 153 as an I/O retry and event 129 as a storage timeout/reset. Those events point to the storage path, which includes the device, connection and controller.

If Windows remains usable, the completed CHKDSK result is in Windows Logs, Application. Filter for the Chkdsk and Wininit sources. This lets you read what the last run found without starting another scan.

Why CHKDSK returns at every startup

Windows tracks whether a volume may contain unfinished changes or corruption using its “dirty bit.” When that flag is set, the next startup can trigger an automatic check. Repeated checks mean the underlying condition still needs attention; suppressing the prompt does not repair the file system. If Windows offers a countdown to skip the check, press a key before it starts. From a working Windows session, an administrator can temporarily exclude a stable volume with chkntfs /x C:. Substitute the correct drive letter and include any existing exclusions in the same command. After preserving the files, chkntfs /d restores the default checking policy for all volumes.

On a stable drive, an administrator can inspect the flag with fsutil dirty query C:, replacing C: with the intended volume. Microsoft's dirty-bit reference distinguishes this query from the command that sets the flag. Querying it is for understanding why the check was scheduled, not a reason to run another repair before copying the files.

Get help with data recovery

Bring your device to our Lincroft lab during opening hours, or mail it in.

How to mail your device

The $100 diagnostic is non-refundable once diagnostics start, and donor drives or replacement parts you approve are non-refundable once ordered, even if recovery fails.

Published by MDrepairs

MDrepairs is a data recovery and electronics repair lab in Lincroft, New Jersey. Our founder and lead technician Joseph Montanti films real cases for our YouTube channel, showing tools, repair steps and recovery outcomes.

Keep reading

Data recovery services

Devices recovered at our lab

These rows come from BenchKey, the ticket system we built for our own shop. Customer names, contact details and descriptions are left out.

Dust and specks visible across a hard drive platter
Dust and specks lie across the exposed platter under green light at our lab in Lincroft.
4.6 · 345 Google reviews$100 diagnostic, credited toward a successful recovery
Start your mail-in recovery