Hard Drive Clicking, Freezing, or Not Showing Up: What To Do Before You Lose Your Data

If your hard drive is clicking, stop reading and unplug it. Everything else here can wait ninety seconds.

That sounds dramatic. Here is why it isn’t. A clicking drive is usually one whose read/write heads are moving where they shouldn’t be, and every extra minute of power is another few thousand passes over the surface holding your files. Data that was recoverable this morning can be genuinely gone by dinner.

If your drive is freezing, running slow, or not showing up at all, you have more time. Not unlimited time, and what you do in the next hour matters more than most people expect. This guide covers what each symptom of a failing hard drive actually means, what to do first, and which piece of very common internet advice is the fastest way to turn a recoverable drive into an unrecoverable one.

We repair and recover drives at our bench in Zanesville, so most of what follows comes from opening the things up and seeing how they failed.

Key takeaways

Clicking or grinding means stop. Power the drive down immediately and do not power it back on to “check if it’s still doing it.”
The single most damaging thing you can do to a mechanically failing drive is run recovery software or CHKDSK on it. Both keep the drive working for hours and can write over the data you’re trying to save.
A drive that won’t show up is often better news than a drive that clicks. The problem is sometimes the enclosure, cable, or port rather than the drive itself.
SMART health checks are not a smoke alarm. Plenty of drives fail with no warning at all, so a clean SMART report does not mean you’re fine.
SSDs give you no audio warning and are usually harder to recover than spinning drives, so backups matter more, not less.
If the data matters and the drive is making noise, the cheapest path is almost always to stop touching it and get it evaluated.

First, the thirty second triage
What clicking, grinding, and beeping actually mean
Freezing, crawling, and blue screens
When the drive doesn’t show up at all
The advice that costs people their data
Why a SMART check can miss a failing hard drive
SSDs fail differently, and quieter
Do you actually need professional recovery?
If this is a business machine
How to not end up here again
Frequently asked questions

First, the thirty-second triage

Everything depends on which of these you’re dealing with.

The drive makes noise it didn’t used to make. Clicking, grinding, scraping, beeping. Power it off now. Do not run anything on it. Skip to the noise section below.

The drive is silent, but the computer freezes, crawls, or blue screens. You probably have some time. Stop creating new files, and if there’s anything on there you can’t lose, copy that specific folder off first before you do anything else.

The drive doesn’t appear at all. Don’t panic yet. Work through the enclosure and cable checks below before assuming the worst.

The drive was dropped, flooded, or was in a machine that took a hit. Off. Now. Physical trauma plus power is the worst combination there is.

One rule covers all four: the more you use a failing drive, the less of it you get back. That’s the whole principle. Every choice below follows from it.

What clicking, grinding, and beeping actually mean

Inside a traditional hard drive, a stack of magnetic platters spins at 5,400 or 7,200 RPM while read/write heads float above them on a cushion of air a few nanometers thick. To put that in perspective, a particle of cigarette smoke is roughly a hundred times bigger than that gap.

Clicking is usually the head assembly failing to find its position, retreating, and trying again. Over and over. Something is wrong with the heads, the servo data they navigate by, or the surface itself. The click is the sound of the retry.

Grinding or scraping typically means the heads are touching the platter instead of flying over it. That contact physically removes the magnetic layer where your data lives. This is the one where minutes matter most, because a scratch is permanent in a way that almost nothing else in this article is.

Beeping usually means the motor is trying to spin the platters and can’t, often because the heads are stuck to the surface, or the spindle has seized. Sometimes it’s simply not enough power reaching a bus-powered external drive, so it’s worth trying a different port or a powered hub once before assuming failure.

In every one of these cases, the drive is still trying. That’s the problem. Leaving it powered doesn’t help it sort itself out; it just gives it more opportunities to damage the surface. Recovery from a mechanical failure means opening the drive in a filtered-air environment and often transplanting parts from a matching donor, and the amount of surface damage when it arrives is the single biggest factor in whether that works.

Freezing, crawling, and blue screens

This group is less urgent and less clear-cut, which cuts both ways.

A drive developing bad sectors will often stall when it hits one, retry for several seconds, then continue. From the outside, that looks like your whole computer hanging for no reason, especially when opening files or booting. File copies that stop dead partway through are a classic sign.

Here’s the honest part that most articles skip: these symptoms are not proof the drive is dying. Failing RAM, an overheating CPU, a bad SATA cable, a corrupt Windows update, and malware all look similar from the user’s seat. We see plenty of machines where the drive was fine, and something else was the culprit.

What to do anyway: assume it might be the drive, because the cost of being wrong in that direction is much lower. Copy your irreplaceable files off first, starting with the ones you’d be sick about losing. Don’t install new software, don’t run a disk optimizer, and don’t defragment anything. Then get it diagnosed properly rather than guessing.

When the drive doesn’t show up at all

An external drive that vanished is worth a few careful minutes before you conclude anything.

Try a different cable first, since USB cables fail constantly and it costs nothing to rule out. Then a different port, ideally one directly on the machine rather than through a hub. Then a different computer entirely. If it’s a 2.5-inch bus-powered drive, an underpowered port alone can cause it to disappear or click.

If the drive spins up and feels normal but still doesn’t mount, the failure may be in the enclosure’s bridge board rather than the drive. That’s the good outcome. One important caveat, though: on many modern external drives, the encryption is handled by that bridge board, so pulling the bare drive out and connecting it directly can leave you with a disk full of unreadable data. If you get to that point, it’s worth asking us first before you start disassembling.

For internal drives, where the drive appears tells you a lot. Visible in Disk Management but with no drive letter or showing as RAW usually points to a partition or file system problem, which is often the most recoverable situation of all. Not visible in the BIOS at all points to hardware.

Whatever you do, if a drive shows up asking to be formatted or initialized, say no. Those prompts are how a fixable problem becomes a much harder one.

The advice that costs people their data

If you search these symptoms, most of what you’ll find recommends running CHKDSK, downloading a recovery tool, or both. That advice is fine for one specific situation and genuinely destructive in another, and almost nobody makes the distinction.

CHKDSK writes to the disk. It isn’t an inspection; it’s a repair tool. On a drive with failing sectors, it will hammer them with read retries for hours and then start moving and deleting data it decides is unusable, which is where those mystery FOUND.000 folders come from. If the file system was your only problem, you may have just made it worse. If the drive was mechanically failing, you left it running under load for the exact hours you couldn’t afford.

Recovery software reads the entire surface. On a healthy drive with a deleted partition, that’s exactly right. On a clicking drive, you’ve just asked damaged heads to make a full pass over every platter. Some tools also default to saving recovered files back to the same drive, which overwrites the thing you’re recovering.

So here’s the split worth remembering. If the drive sounds completely normal and the problem is logical, meaning a deleted file, a formatted partition, or a corrupted partition table, then recovery software is a reasonable first move. Make an image of the drive first, work from the image, and never write anything back to the source. If the drive is making any unusual noise at all, software is the wrong tool and running it will cost you options.

Two more worth naming. The freezer trick was folklore even when it occasionally helped on drives from twenty years ago, and on modern hardware you’re introducing condensation onto the platters and thermal stress onto parts with very tight tolerances. It regularly turns recoverable drives into scrap. And opening the drive yourself in a normal room lets in particles far larger than the gap the heads fly in, which causes exactly the head crash you were trying to avoid. Once a drive has been opened, some labs will charge more or decline the job.

Why a SMART check can miss a failing hard drive

Most drives carry self-monitoring called SMART, and plenty of free tools will read it and give you a reassuring green checkmark. Treat that checkmark with real suspicion.

Google studied roughly 100,000 drives in its own data centers and published the results in Failure Trends in a Large Disk Drive Population. Certain SMART indicators did correlate strongly with failure, and after a first scan error, a drive was dramatically more likely to fail within 60 days. But the paper’s conclusion is the part worth remembering: models built on SMART data alone aren’t reliable for predicting whether an individual drive will fail, because a large share of the drives that died had shown no SMART warning whatsoever.

The practical translation is simple. A bad SMART reading is a serious warning you should act on immediately. A good SMART reading means very little. Your ears and your own observations are better evidence than the checkmark.

Worth keeping in perspective alongside that: drives are not fragile by nature. Backblaze publishes failure data on its entire fleet every year, and across 344,196 drives in its 2025 report, the annualized failure rate was 1.36%. Most drives have a long, boring life. The trouble is that the ones that fail rarely announce it on a schedule.

SSDs fail differently, and more quietly

Nearly every article on this topic quietly assumes a spinning drive, which leaves SSD owners with advice that doesn’t apply. There are no moving parts, so there is no clicking, no grinding, no warning sound of any kind.

What you see instead: files that suddenly become read-only, the whole drive dropping into read-only mode, the disk disappearing from the BIOS and coming back after a reboot, or a machine that boots fine one day and not at all the next. SSDs often go from working to completely unresponsive with very little in between, because when the controller fails the drive stops presenting itself at all.

The uncomfortable part is that SSD recovery is generally harder and less successful than recovery from a spinning drive. Data is spread across chips by wear leveling rather than sitting in one place, most consumer SSDs encrypt at the controller level, and TRIM can permanently clear deleted data within seconds rather than leaving it there to be found. Recovery is often still possible, but the odds are worse, and the window is shorter.

The advice stays the same. Stop using it, don’t run tools on it, and get it looked at. And if you’re running an SSD as your only copy of anything important, treat that as the actual risk rather than the drive’s age.

Do you actually need professional recovery?

Not always, and we’d rather tell you that up front than after you’ve paid for a diagnostic.

You probably don’t need it if the drive sounds normal, still mounts, and you can copy your files off. Do that today, buy a replacement, and move on. You also don’t need it if the data is genuinely replaceable. A drive full of reinstallable games and files that are already synced to a cloud account is a drive you can simply replace.

You probably do need it if the drive is making noise, if it’s not detected at all, if it was dropped or exposed to water, if it’s part of a RAID array, or if the data is something you cannot rebuild: family photos, a business’s financial records, work with no other copy. In those cases, the first attempt is the one most likely to succeed, and every DIY attempt in between reduces what a professional can get back.

Cost depends heavily on whether the failure is logical or mechanical, and mechanical jobs needing donor parts sit at the top of the range. We’ve broken down what data recovery actually costs in a separate article rather than guessing at a number here. Any shop that quotes you a firm price over the phone without seeing the drive is guessing too.

Distance isn’t necessarily a problem either. Drives travel well when they’re packed properly, and our mail-in service exists for exactly this reason.

If this is a business machine

A failed drive at a business is a different problem, because the clock is running on more than the data.

Work out roughly what an hour of downtime costs you before you decide how to handle it, since that number usually makes the decision obvious. Then tell your staff to stop touching it, and call whoever handles your business machines. The most expensive recoveries we see are the ones where a well-meaning employee spent a day trying tools they found online before anyone called us.

If it’s a server or a RAID array, this matters even more. Do not rebuild a degraded array and do not reinitialize it. A rebuild puts every remaining drive under sustained load at precisely the moment they’re least able to handle it, and if a second drive drops during the rebuild you can lose the whole set. Note which drive failed and in what order, then leave it alone.

Write down what happened and when, especially if the machine held customer records, financial data, or anything covered by regulations you’re subject to. If that’s your situation, loop in whoever handles your compliance or legal side early rather than after the fact. We can speak to what happened to the hardware, but the reporting obligations are a conversation for your own advisors.

How to not end up here again

The version that actually works is the 3-2-1 rule, which CISA points small businesses to as the baseline: three copies of anything you care about, on two different kinds of storage, with one of them somewhere else entirely. The offsite copy is the one people skip and the one that saves you when the problem is theft, fire, or ransomware rather than a dead drive.

An external drive sitting next to your computer is one copy on one type of media in one building. It’s better than nothing, and it isn’t a backup plan.

Test it once. Actually restore a file from wherever you think your backup lives. Backups that were silently failing for eight months are common enough that we’ve stopped being surprised, and the moment you discover it should not be the moment you need it.

The short version

A drive that’s making noise is on a clock, and every minute of power spent is part of what’s left. A drive that’s silent but misbehaving gives you a real chance to save your own files if you act before you experiment. And the advice you’ll find first in search results is often the advice most likely to close the door on a professional recovery.

If you’re not sure which situation you’re in, that uncertainty is itself a reason to stop and ask. Powering a failing hard drive down costs you nothing. Powering it up one more time to check occasionally costs everything.

If your drive is clicking, not mounting, or you just want someone to look at it before you make a decision, we’re at 2561 Linden Ave in Zanesville, and we work with customers across southeastern Ohio and by mail. Get in touch about data recovery, book a drop-off, or call us at 740-319-9009.

Frequently asked questions

Is a clicking hard drive always dead?

No, but it’s always serious, and it’s the symptom where your response matters most. Clicking usually points to a problem with the heads or the servo information they use to position themselves, which is a mechanical failure that needs the drive opened in a controlled environment. Plenty of clicking drives are recovered successfully. The ones that aren’t are usually the ones that stayed powered on for hours after the noise started.

Can I recover data from a failing hard drive myself?

Sometimes, and it depends entirely on how it’s failing. If the drive sounds normal, still mounts, and the problem is a deleted file or a damaged partition, DIY recovery software is a fair option as long as you image the drive first and never save recovered files back to the source. If there’s any unusual noise, or the drive isn’t detected at all, DIY attempts tend to reduce what a professional can recover afterward.

Should I run CHKDSK on a drive that’s failing?

Not before you know why it’s failing. CHKDSK repairs file systems by writing changes to the disk, and on a drive with hardware problems it can run for hours under load and then remove data it judges to be corrupt. If your files matter, copy what you can off first, and get the drive diagnosed before running any repair tool on it.

Does putting a hard drive in the freezer work?

No. It’s a persistent myth from an era of very different hardware. Cooling a modern drive that far introduces condensation onto surfaces that must stay perfectly clean and stresses components built to tight tolerances. We’ve received drives that were probably recoverable before someone froze them and were not afterward.

How long do hard drives usually last?

There’s no fixed lifespan, and averages are less useful than people want them to be. Backblaze’s 2025 fleet data put the annualized failure rate at 1.36% across more than 344,000 drives, which tells you most drives are fine most of the time. It doesn’t tell you anything about yours. Age raises the odds, but drives fail in year one, and drives run fine for a decade, which is why a backup is a better plan than a replacement schedule.

My external drive isn’t showing up. Is the drive dead?

Not necessarily, and this is the symptom with the best odds. Try a different cable, a different port, and a different computer before concluding anything, because the enclosure’s circuit board and the cable both fail more often than people expect. If it spins up and feels normal but won’t mount, the drive itself may be fine. Be careful about removing it from its enclosure, though, since many external drives encrypt through that board and the bare drive alone can be unreadable.