Last updated:
SMART monitoring reads health information reported by a hard drive or SSD. It can reveal warnings, temperatures and error trends, but it cannot predict every failure or guarantee that your files are safe. Keep an independent backup even when a utility reports “Good” or “Passed.”
Tools to start with
| Tool | Best starting use | Important limit |
|---|---|---|
| CrystalDiskInfo | A Windows graphical view of drive health and temperature | USB, RAID and controller support depends on the device |
| smartmontools / smartctl | Detailed command-line health reports on supported platforms | Choose the correct device and supported controller interface |
| The drive manufacturer’s utility | Model-specific diagnostics, supported firmware information and warranty checks | Features can be restricted to that maker’s supported drives |
Download CrystalDiskInfo from Crystal Dew World or smartmontools from its official project. Do not confuse CrystalDiskInfo with CrystalDiskMark: the latter benchmarks performance rather than providing the same drive-health overview.
Check the correct disk
- Identify the physical drive by model, capacity and serial number. A drive letter identifies a volume, not necessarily a separate physical disk.
- Record a baseline health report and temperature.
- Review the manufacturer’s meaning for a warning or attribute rather than comparing raw numbers from different brands.
- Recheck trends when symptoms change; one snapshot cannot establish a remaining lifespan.
A read-only smartctl example
smartctl --scan-open
smartctl -a /dev/sdaThe first command discovers accessible devices. Replace /dev/sda with the exact device reported on your system; Windows and NVMe device names can differ. Administrative permissions may be required. Do not copy an example device name blindly or start a write/erase test.
Read the warning in context
- HDD reallocated or pending sectors: investigate growth and symptoms and secure your data.
- Interface CRC errors: check cables and connections; a historical count alone does not prove media failure.
- SSD wear indicators: their interpretation is model-specific and they are not an exact failure countdown.
- NVMe warnings: review critical-warning flags, spare capacity, percentage used, temperature and media errors using the drive’s documentation.
A percentage-used field can represent endurance consumed rather than “percentage healthy.” An unsafe-shutdown count records an event; it does not establish that the drive is already damaged.
When a drive does not appear
A USB enclosure, RAID controller or storage bridge may not pass through every command. Check the utility’s supported-controller documentation. Failure to read SMART through an enclosure is not by itself a failing-drive diagnosis. Conversely, a readable “Passed” report does not exclude a cable, controller, filesystem or sudden electronics problem.
Symptoms matter more than a green badge
If a drive drops offline, makes unusual mechanical noises or struggles to read files, secure important data before stress-testing it. Use the maker’s supported diagnostic route when appropriate. For slow sustained writes, our SMR model-identification guide explains another possible cause. If the whole PC locks up, the Windows freezing checklist covers other components.
Temperature concerns can also warrant checking SSD cooling and clearance. Keep monitoring, performance testing and backups as separate jobs; a monitoring utility does not maintain or repair a drive simply by staying open.