How to Check SSD Health and Read SMART Data on Windows

Hard drives usually warn you before they die. They click, they grind, they slow to a crawl, and you get a few days of dread to back things up. Solid-state drives are not so polite. An SSD can run perfectly on Monday and refuse to enumerate on Tuesday, taking your files with it. The good news is that every modern SSD keeps a running diagnostic log of its own health called SMART, and on Windows you can read it in about two minutes. Learning to interpret a handful of those values is the difference between replacing a drive on your schedule and recovering one on an emergency timeline.

What SMART Actually Tracks

SMART stands for Self-Monitoring, Analysis, and Reporting Technology. It is a set of counters and thresholds the drive maintains internally. On a spinning hard drive, SMART tracks things like reallocated sectors and spin-up time. On an SSD, the attributes that matter are different because the failure modes are different. NAND flash wears out with writes, controllers manage spare blocks, and the firmware tracks how much endurance is left.

The catch is that SMART attribute IDs are only loosely standardized. Two SSDs from different vendors might report wear in completely different attributes, and some vendors rename or remap them. That is why you want a tool that interprets the raw numbers rather than trying to read hex values yourself.

The Fastest Way to Read It: CrystalDiskInfo

For most people, the right tool is CrystalDiskInfo. It is free, open source, and reads SMART data for SATA and NVMe drives. Download the standard installer (or the portable zip if you would rather not install anything), run it, and it shows every drive in your system with a big color-coded health box: blue for good, yellow for caution, red for bad.

Do not stop at the color, though. The color is a rollup, and CrystalDiskInfo is conservative — a single reallocated sector can flip a drive to yellow even when it has years of life left. The useful information is in the attribute table below the health box.

The Attributes That Matter on an SSD

Focus on these:

• Percentage Used / Wear Leveling Count / Media Wearout Indicator. This is the headline number. NVMe drives report a clean “Percentage Used” value in their health log — 0% is new, 100% means the drive has consumed its rated write endurance. Note that 100% does not mean instant death; it means the warranty endurance is spent and you should be planning replacement. SATA SSDs often express the same idea as a “Wear Leveling Count” that counts down from 100.
• Total Host Writes / Data Units Written. How much data has been written to the drive over its life, usually in gigabytes or terabytes. Compare this to the drive’s rated TBW (terabytes written) from the spec sheet. A 1 TB consumer SSD is often rated for 600 TBW. If you are at 50 TBW, you have decades of normal use left.
• Available Spare (NVMe). The percentage of spare blocks remaining. New drives sit at 100%. There is also an “Available Spare Threshold,” typically 10%. When available spare drops below that threshold, the drive is genuinely in trouble.
• Reallocated Sectors / Reallocated Event Count. A few are normal over a drive’s life. A number that climbs week over week is a warning sign.
• Uncorrectable Errors / CRC Errors. These should be zero or very near it. Climbing CRC errors on a SATA SSD frequently mean a bad cable or port, not a dying drive — swap the SATA cable before condemning the disk.
• Power-On Hours and Power Cycle Count. Context, not health. Useful for spotting a “new” drive that has actually been running for three years.

Reading NVMe Health Without Extra Software

If you cannot install anything — a locked-down work machine, for example — Windows can read NVMe health from the command line. Open PowerShell as administrator and run:

Get-PhysicalDisk | Get-StorageReliabilityCounter | Format-List

This returns wear, temperature, read/write error counts, and power-on hours for each physical disk. It is less friendly than CrystalDiskInfo and the wear field is not always populated by every controller, but it works on a stock Windows install with no downloads.

You can also check the basic Windows verdict with:

Get-PhysicalDisk | Select-Object FriendlyName, MediaType, HealthStatus, OperationalStatus

HealthStatus of “Healthy” is reassuring but coarse — it only flips to “Warning” or “Unhealthy” when the drive is already reporting predicted failure. Treat it as a smoke alarm, not a fuel gauge.

Vendor Tools Add One Thing Worth Having

Samsung Magician, Crucial Storage Executive, WD Dashboard, and similar vendor utilities read the same SMART data, but they add two genuinely useful features. First, firmware updates — SSD firmware bugs are real and have caused data loss in the field, so a one-click firmware check is worth running once. Second, they often expose the drive’s rated endurance and an estimated remaining lifespan in plain language, which saves you the TBW math. If you have a single-brand drive, install the matching tool, check for firmware, then you can usually uninstall it.

How to Interpret What You See

Put the numbers together and the picture is usually clear:

• Healthy drive: Percentage Used in single or low double digits, Available Spare at or near 100%, zero reallocated sectors, zero uncorrectable errors, host writes well under the rated TBW. Keep using it. Re-check every few months.
• Aging but fine: Percentage Used climbing past 70–80%, host writes approaching TBW, spare still high. Start budgeting for a replacement and make sure your backups are current. It is not an emergency.
• Replace soon: Available Spare below threshold, reallocated or uncorrectable errors climbing, or the drive throwing read errors you can feel as freezes. Back up now, today, before you do anything else, then clone or migrate.

The single most important habit: if a drive ever flips from green to yellow or red, do not wait to investigate. Copy your irreplaceable files off first, then dig into which attribute changed. SMART is a prediction, not a guarantee, and the safe move is always to assume the prediction is correct until you have your data somewhere else.

A Word on Temperature and Performance

While you have the tool open, glance at the drive temperature. NVMe drives can thermally throttle above roughly 70°C, which shows up as sudden performance drops during large transfers. If a fast drive feels slow under sustained load and the temperature is spiking, the fix is a heatsink or better case airflow, not a new drive. That is a cooling problem masquerading as a health problem — and SMART will tell you the difference if you look.

Checking SSD health is a five-minute habit that quietly pays for itself the first time it catches a drive on the way down. Run CrystalDiskInfo once a quarter, keep an eye on Percentage Used and Available Spare, and let the drive tell you when its time is coming — on your schedule, not its own.