How to Check SSD Health: SMART Data, Free Tools & Warning Signs
Quick Answer
On Windows, download the free CrystalDiskInfo, open it, and read the color-coded health status: blue is Good, yellow is Caution, red is Bad. The two numbers that matter most are Reallocated Sectors Count (should be zero) and Wear Leveling Count (should be well above the manufacturer threshold). On macOS, run sudo smartctl -a disk0 in Terminal (install smartmontools via Homebrew first). On Linux, use sudo smartctl -a /dev/sda.
SSDs fail silently. Unlike hard drives, they rarely make warning sounds or show obvious slowdowns before data loss becomes imminent. The SSD's built-in SMART self-monitoring system logs wear metrics continuously, and reading those logs with the right tool gives you weeks of advance warning — enough time to back up and replace before anything is lost. This guide covers every method across Windows, macOS, and Linux, explains which SMART attributes actually matter, and shows you how to calculate remaining lifespan from TBW data.
What Is SSD SMART Data?
SMART (Self-Monitoring, Analysis and Reporting Technology) is a firmware-level logging system built into every modern SSD. The drive tracks dozens of internal metrics — cell erase counts, error correction rates, temperature history, power-on hours — and stores them in an onboard register. Reading that register with diagnostic software lets you see the drive's health state without running any intrusive tests.
SMART was originally designed for spinning hard drives, but SSD manufacturers added flash-specific attributes covering NAND wear, program/erase cycles, and media error counts. The exact attribute set varies by brand, but several are universally important.
Key SMART Attributes to Check on Any SSD
| Attribute | ID | What to watch for |
|---|---|---|
| Reallocated Sectors Count | 05 | Must be 0 or near 0. Any upward trend means NAND cells are failing and spares are being used. |
| Wear Leveling Count | 177 / E6 | Starts at 100 (or 253 on some drives) and drops with wear. Below 20 is serious. |
| Total Host Writes | F1 / 241 | Cumulative data written in GiB. Compare to the drive's rated TBW specification. |
| Uncorrectable Error Count | BB / 01 | Any non-zero value means data that the ECC engine could not fix. Back up immediately. |
| Power-On Hours | 09 | Context only — age matters less than wear for SSDs, but 40,000+ hours warrants closer monitoring. |
For NVMe drives, also check the Percentage Used field in the NVMe SMART log — it is reported separately from SATA SMART attributes and directly reflects the fraction of rated TBW consumed.
How to Check SSD Health on Windows
CrystalDiskInfo (Recommended — Free)
- Download CrystalDiskInfo from crystalmark.info (use the Standard or Portable edition).
- Run the application. Each connected drive appears as a tab.
- The large colored indicator shows the overall status: Blue = Good, Yellow = Caution, Red = Bad.
- Click any attribute in the list to see its current value, worst-ever value, and the threshold that triggers a warning.
A Caution or Bad status on any attribute warrants a full backup before doing anything else.
Manufacturer Diagnostic Tools
- Samsung SSDs: Samsung Magician — shows health score, remaining TBW, and drive temperature
- WD / SanDisk: WD Dashboard
- Crucial / Micron: Crucial Storage Executive
- Kingston: Kingston SSD Manager
- Seagate: SeaTools for Windows
Manufacturer tools access proprietary health data beyond standard SMART and tend to be more accurate for their specific drives. If you know your SSD brand, the manufacturer's tool is the better first choice.
Quick Command-Line Check
wmic diskdrive get model,statusReturns "OK" or "Pred Fail" per drive. Limited to pass/fail — use it for a fast sanity check when you do not want to install software.
How to Check SSD Health on macOS
System Information (Basic)
- Hold Option and click the Apple menu → System Information.
- Click NVMExpress (for NVMe SSDs) or SATA/SATA Express.
- Select your drive and look for SMART Status: Verified. "Failing" means replace immediately.
Terminal with smartctl (Full Attribute Access)
brew install smartmontools
sudo smartctl -a disk0Replace disk0 with your drive identifier. Run diskutil list to find the correct path. For NVMe drives on newer Macs, add -d nvme.
How to Check SSD Health on Linux
sudo apt install smartmontools # Debian/Ubuntu
sudo smartctl -a /dev/sda # SATA SSD
sudo nvme smart-log /dev/nvme0 # NVMe SSD (requires nvme-cli)For a pass/fail summary only: sudo smartctl -H /dev/sda. Review the full attribute list even on a PASSED result — individual Caution values appear in the detailed output but do not fail the overall health test.
Understanding TBW — SSD Rated Lifespan
Manufacturers rate SSDs by TBW (Terabytes Written) — the total data the drive is warranted to accept before NAND wear becomes a concern. To check how much you have used, read the Total Host Writes SMART attribute (in GiB; divide by 1,000 to convert to TBW).
| Drive type | Typical TBW (1 TB) | Expected lifespan (typical home use) |
|---|---|---|
| Budget QLC SATA | 100–150 TBW | 3–5 years |
| Mid-range TLC SATA | 200–400 TBW | 5–8 years |
| High-performance NVMe TLC | 300–600 TBW | 5–10 years |
| Enterprise NVMe | 1,000–5,000+ TBW | 10+ years under heavy load |
Early Warning Signs of SSD Failure
- Sudden unexplained slowdowns on a drive that previously ran at rated speeds
- Files that cannot be opened, copied, or written — applications crashing on disk operations
- The drive disappearing from File Explorer or Device Manager intermittently
- Blue screen errors referencing disk I/O failures (event ID 11, 153 in Windows Event Viewer)
- CrystalDiskInfo or smartctl showing Caution or Bad on any attribute — especially Reallocated Sectors or Uncorrectable Errors
If you see any of these, stop writing data to the drive and back up immediately before running further diagnostics.
FAQs
Can an SSD fail suddenly without any SMART warning?
Yes, though uncommon. Controller IC failures or electrical damage can cause instant data loss with no prior SMART warning. This is why backup discipline matters regardless of health status — SMART predicts gradual wear failures well but cannot predict sudden hardware failures.
Does defragmenting an SSD harm it?
Running traditional defragmentation wastes write cycles and provides no performance benefit — SSDs access all locations at equal speed. Windows 10 and 11 automatically send TRIM commands to SSDs instead. Never run a mechanical defrag tool on an SSD.
How long do SSDs typically last?
Consumer SSDs last 5–10 years under typical home use before reaching their rated TBW. Many drives outlive their TBW specification by 2–3× in real-world conditions. The more useful predictor than age is total data written compared to TBW rating.
Is 100% health normal for a new SSD?
Yes. A new SSD should show 100% (or very close to it) in CrystalDiskInfo, zero Reallocated Sectors, and near-zero wear values. Some drives calibrate slightly below 100% fresh out of the box — values between 95–100% are not concerning.
Should I worry if CrystalDiskInfo shows Caution?
Yes. Back up all important data immediately, then identify which attribute triggered the Caution. Caution on Wear Leveling Count means the drive is aging. Caution on Reallocated Sectors means cells are actively failing. Either way, plan replacement soon.
Summary
Checking SSD health takes under two minutes using CrystalDiskInfo on Windows or smartctl on macOS and Linux. The most critical indicators are Reallocated Sectors Count (must stay at zero) and Wear Leveling Count (must stay well above threshold). Compare Total Host Writes to the drive's rated TBW to estimate remaining lifespan. Any Caution or Bad SMART status is a sign to back up immediately and schedule replacement.