CrystalDiskInfo Guide: How to Check Hard Drive and SSD Health

The essential tool for evaluating used drives | Updated 2026

CrystalDiskInfo is a free, open-source utility that reads SMART data from your hard drives and SSDs and displays it in a clear, color-coded interface. If you're buying used hard drives or used SSDs on eBay, this is the single most important tool you need.

What Is CrystalDiskInfo?

Every hard drive and SSD has a built-in self-monitoring system called SMART (Self-Monitoring, Analysis, and Reporting Technology). The drive tracks dozens of health metrics internally — error counts, temperature history, hours of use, and more. CrystalDiskInfo reads this data and presents it in a way humans can actually understand.

Download

Official site: crystalmark.info/en/software/crystaldiskinfo/
License: Free and open source (MIT)
Platform: Windows (for Linux, use smartctl from the smartmontools package)
Choose "Standard Edition" — the themed editions are the same tool with anime skins

The Health Status Indicator

CrystalDiskInfo shows one of four status badges at the top of each drive:

Status	Color	What It Means
Good	Blue	All SMART attributes are within normal thresholds. The drive is healthy.
Caution	Yellow	One or more attributes have crossed a warning threshold. The drive may be developing issues — investigate which attribute triggered it.
Bad	Red	One or more attributes have crossed a critical threshold. The drive is failing or has failed. Back up data immediately and replace it.
Unknown	Gray	CrystalDiskInfo can't read SMART data. Common with some USB enclosures, RAID controllers, or drives behind hardware RAID cards.

CrystalDiskInfo showing Good (blue) status on a Samsung 970 EVO Plus NVMe SSD with 100% health — Good — Samsung 970 EVO Plus NVMe SSD. Health at 100%, all SMART attributes normal. Note the NVMe-specific attributes like Available Spare and Percentage Used.

CrystalDiskInfo showing Caution (yellow) status on a Toshiba HDD with reallocated sectors — Caution — Toshiba 1TB HDD. The yellow dot on Reallocated Sectors Count (raw value: 28) triggered the warning. This drive is developing bad sectors but hasn't crossed the critical threshold yet.

CrystalDiskInfo showing Bad (red) status on a WD HDD with multiple failing attributes — Bad — WD 750GB HDD. Reallocated Sectors at 140, plus non-zero Current Pending and Uncorrectable Sectors. This drive is actively failing — back up and replace immediately.

"Good" Doesn't Mean Perfect

The "Good" status only means no attribute has crossed its manufacturer-set threshold. A drive can show "Good" while still having concerning values — for example, a non-zero Reallocated Sector Count that hasn't yet hit the threshold. Always read the individual attributes, not just the badge.

The Top Info Bar

Below the health badge, CrystalDiskInfo shows key summary info:

Field	What It Shows	What to Check
Serial Number	Unique drive identifier	Match to listing photos when buying used. Verify it matches between seller's screenshot and yours.
Interface / Transfer Mode	SATA, NVMe, or USB + link speed	SATA III should show "SATA/600". USB enclosures may hide some SMART data.
Temperature	Current drive temperature	HDDs: under 45°C is ideal. SSDs: under 50°C idle, under 70°C under load. NVMe can run hotter.
Power On Hours	Total hours the drive has been powered on	See our HDD buying guide for what counts are acceptable by drive class.
Total Host Writes	Total data written (SSDs only)	The most important metric for used SSDs. See our SSD buying guide for TBW evaluation.

SMART Attributes: The Full Table

The main panel shows all SMART attributes in a table. Here's what the columns mean:

Column	Meaning
ID	Hex code identifying the attribute. Standardized for common attributes, but vendors can add custom ones.
Current	The current normalized value (usually 0-100 or 0-253). Higher is generally better. This is a vendor-specific scale.
Worst	The lowest normalized value this attribute has ever reached. If Worst is close to Threshold, the drive has been near failure.
Threshold	The failure threshold set by the manufacturer. If Current drops below this, the drive is considered failing.
Raw Value	The actual raw count. This is what you should read. The normalized values are often confusing — raw values tell the real story (e.g., "5" reallocated sectors, "23,456" power-on hours).

Focus on Raw Values

The Current/Worst/Threshold columns use a normalized scale that varies between manufacturers and is often confusing. A "Current" value of 100 might be perfect for one attribute and meaningless for another. Always read the Raw Value column — it gives you the actual count in human-readable numbers.

Critical SMART Attributes for HDDs

For hard drives, these are the attributes that matter most when evaluating a used drive:

ID	Attribute	Good Value	Bad Value
05	Reallocated Sector Count	0	Any non-zero and rising — drive has bad sectors
09	Power-On Hours	< 40,000 for enterprise	> 60,000 or exactly 0 (wiped)
0A	Spin Retry Count	0	Any non-zero — motor problems
C5	Current Pending Sector Count	0	Any non-zero — active bad sectors
C6	Uncorrectable Sector Count	0	Any non-zero — unrecoverable read errors
C7	UltraDMA CRC Error Count	0	Low count may be cable issue, thousands = problem
C2	Temperature	< 40°C	> 45°C needs better airflow

Critical SMART Attributes for SSDs

SSDs track different attributes. NVMe drives use a different SMART specification than SATA SSDs, but CrystalDiskInfo normalizes them:

Attribute	Good Value	Bad Value
Total Host Writes	< 50% of TBW rating	> 80% of TBW rating
Percentage Used	0-50%	> 80%, or 100%+ (exceeded rated life)
Available Spare	100%	< 10% — running out of spare NAND
Media Errors	0	Any non-zero — NAND failing
Unsafe Shutdown Count	Low	Very high count — many unexpected power losses
Temperature Warning Count	0	High count — drive overheated frequently

Linux Alternative: smartctl

If you're on Linux (or prefer the command line), smartctl from the smartmontools package gives you the same SMART data:

Quick smartctl Commands

sudo apt install smartmontools — install on Debian/Ubuntu
sudo smartctl -a /dev/sda — show all SMART data for a SATA drive
sudo smartctl -a /dev/nvme0n1 — show all SMART data for an NVMe drive
sudo smartctl -t long /dev/sda — start a long (extended) self-test
sudo smartctl -l selftest /dev/sda — view self-test results
sudo smartctl -H /dev/sda — quick health check (PASSED/FAILED)

For detailed smartctl documentation, see the smartmontools project.

Common Questions

Can SMART data be faked or wiped?

Yes. The telltale sign: a manufacture date years ago but Power-On Hours at exactly 0. Always match the label date to SMART data when buying used.

My drive shows "Caution" — is it dying?

Check which attribute triggered it. A few Reallocated Sectors on an old drive is normal wear. But non-zero Current Pending or Uncorrectable Sectors means active problems — back up and replace.

Why does CrystalDiskInfo show "Unknown" for my drive?

Common with USB enclosures and hardware RAID controllers that don't pass through SMART commands. Connect the drive directly via SATA or M.2.

How often should I check SMART data?

New used drive: daily for a week, weekly for a month, then monthly. Set CrystalDiskInfo to run at startup in the system tray for automatic alerts.