Check disk health on Windows: diagnose and fix it from the command line

What this covers

Strange slowdowns, files that won't open, or error messages about a drive can mean anything from a failing disk to simple file system corruption. This guide walks you from the hardware level up: confirm the drive itself is healthy, scan the file system, and schedule a repair if needed.

Symptom

Check the physical disk's health status

Start at the hardware layer. This shows each drive's health status (Healthy, Warning, Unhealthy) plus reliability counters like temperature, wear, and read errors. If any drive reports Warning or Unhealthy, back up your data immediately — the rest of this guide is about file system issues, not a failing drive. A Healthy result on all drives means you can move on to checking the file system.

$ Get-PhysicalDisk | Select-Object FriendlyName, MediaType, HealthStatus, OperationalStatus, Size
Get-PhysicalDisk | Get-StorageReliabilityCounter | Select-Object DeviceId, Temperature, Wear, ReadErrorsTotal

Scan the volume for file system errors

Even a physically healthy disk can have NTFS file system errors or bad sectors. Run an online scan on the system drive — it checks while Windows is running and fixes nothing, making it a safe diagnostic. A clean result ('no problems found') rules out file system corruption; reported errors mean you should schedule a repair (next step).

$ chkdsk C: /scan

Rule out low free space (optional)

Disk 'health' complaints are sometimes just a nearly full drive causing slowness or errors. This optional check shows used and free gigabytes per drive. If a drive is nearly full, freeing space may resolve the symptoms without any repair needed.

$ Get-PSDrive -PSProvider FileSystem | Format-Table Name, @{n='Used(GB)';e={[math]::Round($_.Used/1GB,1)}}, @{n='Free(GB)';e={[math]::Round($_.Free/1GB,1)}}

Check the dirty bit and schedule a boot-time repair

If the earlier scan found problems, first check whether the volume is marked dirty. Then schedule a boot-time disk check so a full chkdsk repair runs at the next restart, when the drive isn't in use. After restarting, let the check finish — it can take a while on large drives. If the volume isn't dirty, you can skip the repair.

$ chkntfs C:
chkntfs /c C:

$ chkntfs /x C:

How to confirm it worked

After any repair, re-run the physical disk health query — every drive should report Healthy — and the file system scan should complete without reporting errors.

Undo everything

Every command above can be reverted. Run these in order to undo all of it.

$ chkntfs /x C:

Review before running — some steps may not be independent.

Sources

• Get-PhysicalDisk
  Microsoft Learn · vendor-doc · 2026-09-17
  Evidence: Documents HealthStatus and OperationalStatus properties on disk objects.
• Get-StorageReliabilityCounter
  Microsoft Learn · vendor-doc · 2026-09-17
  Evidence: Documents temperature, wear, and error counters piped from Get-PhysicalDisk.
• chkdsk | Microsoft Learn
  Microsoft · vendor-doc · 2026-09-17
  Evidence: Microsoft documents chkdsk /scan as running an online scan on a volume, and /f with /r as locating bad sectors and recovering readable information.
• Check Disk - chkdsk - Windows CMD
  SS64 · technical · 2026-09-17
  Evidence: SS64 explains that chkdsk without switches runs in read-only mode, while /f fixes errors found on the disk.
• Get-PSDrive
  Microsoft Learn · vendor-doc · 2026-09-17
  Evidence: Documents Get-PSDrive -PSProvider FileSystem and the Used/Free byte properties of drive objects.
• chkntfs
  Microsoft Learn · vendor-doc · 2026-09-17
  Evidence: Documents chkntfs volume reporting of the dirty bit and /c, /x scheduling switches.

Last verified: 2026-09-21

Related guides