SMART report explainer: is your drive failing?
In short: paste a SMART report from smartctl or CrystalDiskInfo and this free explainer reads every attribute in plain words. It flags the five attributes Backblaze tied to drive failure — reallocated, pending and uncorrectable sectors, reported errors and command timeouts — and gives you one action: back up now, keep watching, or no signs of failure.
Works with smartctl (Linux, macOS, Windows), CrystalDiskInfo and smartctl for NVMe drives. Nothing you paste leaves your browser.
Nothing to read yet
Paste the whole report, including the table of attributes — not just the summary line.
How to get a SMART report from any drive
Any of these produce text you can paste above. Nothing you paste leaves your browser.
- Windows — CrystalDiskInfo. Pick the drive at the top, then use Edit > Copy. That puts the whole report on the clipboard, including the attribute table.
- Linux and macOS — smartctl. Install smartmontools, then run
sudo smartctl -a /dev/sdafor a SATA drive orsudo smartctl -a /dev/nvme0for an NVMe one. Use-d satif the drive is in a USB enclosure. - A drive in a USB enclosure. Many enclosures hide SMART data entirely. If you get nothing, connect the drive directly to a SATA port or try a different adapter.
- A drive that is already failing. Reading SMART is harmless, but repeated read attempts on a dying drive are not. If it holds the only copy of something important, image it first and investigate the image.
The SMART attributes that actually predict failure
Backblaze, which publishes statistics on a fleet of over 200,000 drives, found one or more of these five above zero in roughly three quarters of its failed drives, and in only about 4% of its working ones. Trust them over the drive's own "PASSED" line.
- SMART 5 — reallocated sector count: sectors that went bad and were swapped for spares.
- SMART 197 — current pending sector count: sectors failing right now that have not been moved yet.
- SMART 198 and 187 — uncorrectable sector count: sectors and reads the drive could not fix at all.
- SMART 188 — command timeout: operations that never completed, often power or cable trouble.
A clean report is not a guarantee: around a quarter of the drives Backblaze saw fail showed none of the five beforehand. SMART catches gradual wear, not a controller that dies suddenly.
From a failing drive to a paid recovery job
If you fix drives for customers, the SMART findings belong on the job. BytePhase data recovery management software keeps the diagnosis, the approval and the handover in one place.
Recovery ticket with the findings
Record the drive and what SMART showed on the job, so the history stays with the customer.Diagnosis report by email, WhatsApp or SMS
Send the customer the diagnosis and get their digital signature approval before you start.Built for computer repair counters
Share the recovered file structure behind an OTP, so only the customer sees it.SMART report FAQs
Self-Monitoring, Analysis and Reporting Technology: a set of counters and measurements every hard drive and SSD keeps about its own health. Reading them tells you what the drive has noticed about itself, which is usually earlier than anything you would notice.
Trust the numbers. The overall health line only turns to “bad” when an attribute crosses the manufacturer’s threshold, and those thresholds are set low enough that a drive with thousands of reallocated sectors can still report good. The sector counters tell you the truth sooner.
It is the warning a computer shows at start-up when the drive reports that one of its own SMART attributes has crossed the manufacturer’s failure threshold — the drive is declaring itself failed, which only happens once damage is well advanced. Copy the data off first, ideally by imaging the drive, then replace it. Paste its SMART report above to see which attribute tripped.
No. Around a quarter of the drives Backblaze saw fail showed none of the five warning attributes beforehand. SMART catches gradual wear, not a controller that dies suddenly or a head crash. A clean report is a reason not to worry, never a reason to skip backups.
No. The parsing runs entirely in your browser; the report is never sent to us or anywhere else. You can disconnect from the internet and the explainer keeps working.
Sources: smartmontools attribute documentation; Backblaze, “What SMART stats tell us about hard drives”. Attribute data checked 15 Oct 2026.
Keep every drive diagnosis on the job
BytePhase keeps the drive, the diagnosis, the customer's approval and the handover on one repair job, so nothing lives in a notebook. 15-day trial.
Last updated: 21 November 2026
