How to Check Hard Drive Health on Windows 11

If your laptop has an NVMe SSD — anything manufactured after 2021 or so probably does — Windows 11 has a built-in health check most people don’t know about. Settings, System, Storage, scroll down to Advanced storage settings, Disks & volumes, click your drive and hit Properties. You’ll see estimated remaining life as a percentage, available spare blocks, and temperature. Anything above 90% remaining life on a drive under two years old is completely normal. Below 50% means start planning a replacement. Below 10% means stop reading this and back up everything right now.

That only works for NVMe though. SATA SSDs and spinning hard drives don’t show that page. For those you need CrystalDiskInfo — free, portable, no installation. Download it, run it, and it immediately shows a colored health box for every drive in your system. Blue means Good. Yellow means Caution. Red means Bad.

A customer brought in an HP Pavilion, three years old, boot time had crept from thirty seconds to over two minutes. She thought maybe malware. I ran CrystalDiskInfo — ten seconds — and it showed yellow. Caution. The drive had 47 reallocated sectors and was running out of spare room. We cloned it to a new SSD that afternoon and she emailed a week later saying the laptop felt brand new. The processor, RAM, screen, keyboard were all fine. The hard drive was the only thing dying, and it had been slowly failing for months while she blamed the whole machine.

CrystalDiskInfo — What to Actually Look At

The colored health box is a starting point but it doesn’t tell the whole story. I’ve seen drives CrystalDiskInfo rated “Good” fail within a month because the specific attributes predicting failure hadn’t crossed the tool’s default thresholds yet.

For spinning hard drives, the single most important number is Reallocated Sector Count. When the drive finds a sector that can’t hold data anymore, it shuffles the data to a spare sector and marks the bad one as dead. Zero on a new drive. Under 5 is normal wear over years of use. Once it creeps past 10-20, the drive is actively degrading — that customer’s HP was at 47. If this number is climbing between checks, stop planning and start backing up.

Current Pending Sector Count tracks sectors the drive suspects are bad but hasn’t confirmed yet. Even one means the drive found something sketchy. More than 5 gets an immediate backup recommendation. Spin Retry Count is a different kind of scary — it means the drive motor needed multiple attempts to spin up the platters. Healthy drives never retry. Drives that retry tend to fail suddenly one morning when the motor just doesn’t start.

Power-On Hours isn’t a failure predictor by itself, but it frames everything else. Consumer drives are rated for 30,000-50,000 hours. At 8 hours a day, that’s 10-17 years. Seeing 40,000 hours alongside even minor yellow flags means the drive earned its retirement.

SSDs fail differently because nothing mechanical wears out. They also don’t need defragmenting — what Windows calls “Optimize” on an SSD is TRIM, not defrag. The number to watch is Percentage Used, sometimes labeled Host Writes or Total NAND Writes. A typical 1TB SSD rated for 600 TBW at 30 GB daily writes would last over 50 years. In practice, consumer SSDs almost never die from write wear — they die from firmware bugs, power loss during writes, or the controller silicon giving up. So low Percentage Used doesn’t guarantee safety the way people assume. Watch Reallocated NAND Blocks for sudden jumps — a few accumulating over years is fine, a bunch appearing in one month isn’t.

Command-Line Quick Check

If you don’t want to install anything, open an admin terminal and run wmic diskdrive get model,status — returns either “OK” or “Pred Fail” for each drive. “OK” means the drive’s firmware hasn’t flagged anything critical. “Pred Fail” means the drive is predicting its own death — back up immediately. The problem with WMIC is that “OK” can be misleading — a drive with 40 reallocated sectors might still report “OK” because it hasn’t crossed its internal failure threshold. WMIC is a fire alarm, not a smoke detector.

Microsoft has been deprecating WMIC since Windows 11 22H2 so the PowerShell replacement is Get-PhysicalDisk | Select-Object FriendlyName, MediaType, HealthStatus, OperationalStatus — gives you the same pass/fail plus whether each drive is SSD or HDD.

For file system integrity specifically, chkdsk C: /f checks and repairs corruption caused by bad sectors, sudden power losses, or buggy software. It’s not a hardware health check — it doesn’t read SMART data — but if chkdsk keeps finding new errors every time you run it, something is corrupting data faster than the file system can track, and the drive is almost certainly the source.

When to Act and What to Watch For

If CrystalDiskInfo shows blue “Good” with zeroes across the board, close the program and forget about it for six months. Just make sure your backup system actually works, because even perfectly healthy drives die from power surges or falls off desks.

Yellow “Caution” is where I stop being relaxed. Reallocated sectors above 20, or an SSD showing significant wear. The drive is degrading faster than its spare capacity can absorb. Back up everything today, not this weekend. Clone to a new SSD within two weeks. Maybe the drive lasts six months, maybe it doesn’t boot on Thursday. Drives at Caution typically have 2-6 weeks before they start losing files or failing to boot.

Red “Bad” or WMIC returning “Pred Fail” is urgent. The drive’s own firmware is predicting death. Stop saving new files to it. Copy irreplaceable files to an external drive right now.

Physical symptoms matter just as much as SMART numbers because SMART only catches about 60-70% of impending failures. The rest — firmware bugs, sudden controller death, power surge damage — don’t register in SMART data at all. New clicking or grinding sounds on an HDD mean heads crashing or bearings struggling, and healthy drives don’t make new noises. Progressively slower boots with no new software installed is a drive reading sectors slower as they degrade. Corrupted files — a Word document that opens as gibberish, a photo with the bottom half grey — means the sector went bad after you saved the file. Random freezes with the disk light pinned solid, or the drive disappearing from Windows entirely and reappearing after restart — check the SATA cable first because a loose cable causes the same symptom, but if the cable is fine, the drive controller is failing intermittently. If SFC /scannow keeps finding new corruption every run, the drive is almost certainly why, and we can run full diagnostics and clone to a new drive remotely before it stops responding entirely.