How to Tell If Your Hard Drive Is Failing: Warning Signs and the SMART Test

A hard drive or SSD that is about to fail almost always leaves signs before the breakdown: unexplained slowdowns, freezes, files that no longer open, clicking noises on a mechanical drive, or errors reported by the SMART monitoring tool. The right reaction is not to repair, but to back up your files immediately.

When a drive dies, what you lose is not settings but your photos, your documents and years of files. Spotting the early signs and checking the drive's health in a few minutes can spare you a real disaster. Here is how, without paid software.

1. A single suspicious sign (slowdowns, noise, unreadable files) justifies backing up right now, before any diagnosis.
2. SMART status can be read for free in a few minutes; a "Caution" or "Bad" alert should be taken seriously.
3. A failing drive cannot be repaired: it is replaced, after copying your data.

1. Slowdowns with no software cause

The disk is slow to respond, the computer freezes for a few seconds when opening a folder, the disk indicator in Task Manager stays at 100% with no demanding program. This symptom has many causes, but a drive struggling to read its data is one of them.

2. Clicking, grinding or odd whirring noises

On a mechanical hard drive, repeated clicks or a grinding sound often point to a read-head or motor problem. If you hear such a noise, stop using the drive as much as possible and copy the essentials without delay. SSDs are silent and do not give this signal.

3. Files that become unreadable or corrupted

A document that no longer opens, a half-displayed photo, an archive flagged as corrupt: if this affects several files for no reason, the disk may have bad sectors.

4. Blue screens or crashes at startup

Recurring blue screens, especially at startup, can come from a struggling disk. Our article on the Windows 11 blue screen explains how to tell this case from a simple faulty driver.

5. A very slow boot or a disappearing disk

A computer that takes several minutes to start, or a secondary drive that vanishes from Explorer and comes back, often signals a read or connection problem. Also consider the cable or USB port before condemning the drive.

6. Errors during a disk check

If the Windows check tool (chkdsk) reports bad sectors or repeated errors, the drive is ageing badly. A one-off error on a file can be fixed; errors that keep coming back are a bad sign.

7. A SMART alert

This is the most objective sign, explained in the next section. It can appear before any visible symptom.

A detail on interpreting the values. SMART attributes show a "raw" value and a "normalised" value that the manufacturer lowers as health degrades. What matters is not the value itself but its trend: a reallocated-sector counter going from 0 to 8 then to 40 within a few weeks is far more worrying than a counter steady at 8 for two years. So record the status once, then re-check from time to time to see the trend.

SMART is a self-monitoring system built into almost every hard drive and SSD. The drive constantly measures its own condition (reallocated sectors, read errors, temperature, wear) and can flag that it is in poor health. It is not infallible: a drive can fail without a SMART alert, and an alert does not mean immediate failure. But it is the best indicator available.

To read it, the simplest method is a free utility such as CrystalDiskInfo: it shows the overall status (Good, Caution, Bad) and the detail of each attribute. Windows also provides a summary health state: in PowerShell, the Get-PhysicalDisk command reports "Healthy", "Warning" or "Unhealthy" for each disk.

• Reallocated sectors: the drive had to replace defective areas. A few are normal, a fast rise is worrying.
• Pending and uncorrectable sectors: areas the drive can no longer read. This is a serious signal.
• Temperature: a drive that is regularly too hot ages faster.
• For an SSD: the wear percentage and the remaining spare level show how much life the drive has left.

If the drive is already badly affected, to the point of not starting or no longer being recognised, caution changes nature. Avoid plugging it in again and again: each attempt can make things worse. A lab data recovery service can sometimes save files, but it generally costs several hundred euros, which makes preventive backup infinitely more cost-effective.

If the drive makes noises, do not shake it, do not put it in the freezer (a persistent myth) and never open a mechanical hard drive: dust alone is enough to destroy it. Copy what you can, starting with the most precious files rather than the largest, because if the drive dies mid-copy you will at least have saved the essentials.

1. Back up first. Copy your important files (photos, documents, mail) to an external drive or online storage. Do it before any check that stresses the drive.
2. Avoid heavy operations: defragmentation, full scans, reinstalling. They make the drive work hard and can hasten failure.
3. Check the cable and port if the drive is external, before concluding it has failed.
4. If the alert is confirmed, replace the drive. A 500 GB SSD generally costs between €40 and €70, a modest expense compared with losing your data.
5. To change drives without reinstalling everything, a cloning tool copies the old drive to the new one, system included.

Keep in mind that a good backup follows the 3-2-1 rule: three copies of your data, on two different media, one of them off-site. A single copy on the computer's own drive is not a backup.

For an SSD, the lifespan quoted by manufacturers is expressed in amount of data written. Office or family use almost never approaches the limit: SSD failures more often come from a controller or firmware fault than from cell wear. That is why an SSD can look perfectly healthy by SMART, then vanish all at once. Here again, backup remains the only real protection.

External drives deserve the same attention. An external drive that is moved often, unplugged without ejecting and dropped now and then lives more dangerously than an internal one. If your only copies of family photos are on such a drive, it is time to make a second copy.

A mechanical hard drive is sensitive to shocks and to wear of its moving parts. It sometimes warns through noise, but also fails without notice. An SSD has no moving parts and resists shocks better, but its wear depends on the volume of data written, and a failure can be sudden. In both cases, SMART monitoring and backups are essential.

A mechanical drive more than four or five years old that slows the computer is a good candidate for replacement with an SSD: the machine regains its snap, and you reduce the risk of failure.

Nyctale asks Windows for the health state of your drives and clearly flags if one is in warning. It cross-checks this with recent crashes and disk activity, to tell a simple software slowdown from a real hardware problem. If the drive is at fault, it tells you frankly: no software repairs a wearing drive, you need to back up and replace it.