Rightek Rescue · Guides · Windows detected a hard disk problem

"Windows detected a hard disk problem"? Here's what to do

This warning covers two very different situations: a fixable filesystem glitch (Windows' index of your files got scrambled) or a drive that's physically failing. The most important thing first: back up anything irreplaceable now, before doing anything else — and we'll be honest with you about which one it is.

Updated 23 August 2026 · written and tested by Rightek IT Services

What you're seeing

"Windows detected a hard disk problem" — back up your files immediately
File-system errors in Event Viewer (Ntfs events), files that won't open, or folders that vanish
Clicking/grinding noises, very slow copies, or the drive disappearing and reappearing — these lean toward hardware failure

Why it happens

Logical (often fixable): a bad shutdown or power loss scrambled the NTFS file table — chkdsk can usually repair this
Physical (not fixable by software): the drive has bad sectors, a failing head, or worn-out flash — SMART data shows reallocated/pending sectors climbing
An SSD near end-of-life, or a laptop drive that took a knock

Read the drive's own verdict first

Windows' warning is a blunt instrument — it fires on a handful of triggers and does not tell you how bad things are. The drive itself keeps far better records, in its S.M.A.R.T. data. The numbers that matter:

Reallocated Sectors Count — sectors that failed and were swapped for spares. A small number that stays put is tolerable. Any number that is climbing means the drive is actively degrading.
Current Pending Sector Count — sectors that failed to read and have not yet been dealt with. Anything above zero deserves attention; these are where unreadable files come from.
Reported Uncorrectable Errors — reads the drive could not fix. Above zero is a bad sign.
Percentage Used / Media Wearout on SSDs — how much of the write endurance has been consumed.

One reading is a snapshot. Two readings a week apart tell you the direction, which matters more. A drive with a stable count of a few reallocated sectors may run for years; one that added twenty this week will not.

What to copy off, and in what order

If the numbers are bad, treat the machine as having limited reads left and plan accordingly. Copy the irreplaceable things first, not the largest folder:

Photos and videos — the genuinely irreplaceable category
Documents, and anything work-related
C:\Users\Public, which people forget
Browser profiles, if bookmarks and saved logins matter
Everything else, if the drive is still cooperating

Copy to a different physical drive, not another partition of the same one. And check the copies opened properly before you rely on them — a struggling drive can produce files that copy without error and are truncated.

Should you run chkdsk?

It depends entirely on which problem you have, and getting it the wrong way round is costly:

If S.M.A.R.T. is clean and this looks like filesystem corruption — after a power cut, say — then yes. chkdsk C: /f repairs the structure and is quick.
If S.M.A.R.T. shows pending or reallocated sectors climbing, then no, not yet. chkdsk /r in particular reads and rewrites every sector for hours, which is sustained stress on media that is already failing. Copy your data off first. Run repairs afterwards, if at all.

The rule of thumb: repair the filesystem on a healthy disk, rescue the data from an unhealthy one. Doing it the other way round is the single most common way people lose files they could have kept.

How long has it got?

Nobody can tell you, and be sceptical of anyone who claims to. Drives fail suddenly and drives limp along for years, from identical-looking data. What is predictable is the shape of the risk: the cost of acting early is one external drive and an evening; the cost of acting late is everything on it. That asymmetry is the whole argument, and it does not require knowing the timeline.

⚡ The honest way: Rightek Rescue tells you which it is — free

Rescue reads the drive's own SMART health data and file-system errors: if it's a logical glitch it can run the repair, and if the drive shows warning signs of physical failure it flags them — it won't pretend software can fix dying hardware. That honesty is the point: you find out before you spend money.

If the drive is failing and the data matters, stop using it (every power-on risks more) and let the people at Rightek IT Services recover it — professional drive & data recovery, with an honest quote up front, not a scare tactic.

Get Rightek Rescue → Data recovery service

What you can do yourself

Back up first. Copy anything irreplaceable to another drive or the cloud before running any repair.
For a logical glitch: chkdsk /f on the drive (Windows will schedule it for the next reboot).
Check the drive's SMART status with any free tool — rising "reallocated" or "pending" sectors mean the hardware is going.
If it's clicking, grinding, or vanishing: don't run repairs — power it down and get it recovered professionally.

FAQ

Can software fix a failing hard drive?

No - and anyone who tells you otherwise is being dishonest. Software can fix a scrambled file system, but a physically failing drive needs the data recovered off it, not "repaired." Rescue tells you which situation you're in for free.

Is it safe to keep using the drive?

If the drive is failing, no - every power-on can lose more data. Back up what you can immediately and stop using it if it's making noises or dropping out.

My files won't open but the drive seems fine - can I get them back?

Often yes. If it's a logical (filesystem) problem, Rescue's file recovery can pull your files off even when Windows can't read them normally.

Which Windows versions are supported?

Windows 10 and 11, and Rescue runs from a USB so it works even if Windows won't start.

What does S.M.A.R.T. actually tell me?

It is the drive's own health log. The useful figures are Reallocated Sectors Count, Current Pending Sector Count and Reported Uncorrectable Errors. A small, stable number of reallocated sectors is tolerable; a count that is climbing week to week means the drive is actively degrading.

Should I run chkdsk on a drive Windows is warning about?

Only if the drive is otherwise healthy. On a sound disk with filesystem corruption, chkdsk /f is quick and fixes it. On a disk with climbing pending or reallocated sectors, chkdsk /r reads and rewrites every sector for hours, which is sustained stress on failing media. Copy the data off first.

Can the warning be a false alarm?

Yes, and it happens. The warning fires on a limited set of triggers and can be prompted by a cable fault, a controller problem or a one-off read error rather than a dying disk. That is why reading the S.M.A.R.T. figures matters — they distinguish a genuine problem from a scare.

How long will the drive last?

Nobody can say, and treat confident answers with suspicion — drives fail suddenly and drives limp on for years from identical-looking data. What is certain is the asymmetry: acting early costs an external drive and an evening, acting late costs everything on it.