Why Your PC Keeps Restarting on Its Own

You're in the middle of something and the screen goes black. The fans stop, silence for a second, and then the PC boots back up like nothing happened. No warning, no error message, just a restart you didn't ask for.

I get calls about this regularly. People put up with it for weeks because the PC still works fine between restarts, and they assume it's Windows being Windows. Sometimes it is. But random restarts can also be the first sign of hardware that's on the way out, and ignoring it long enough can turn a cheap fix into a dead machine.

Here's what usually causes it.

It's overheating

This is the number one cause. Probably half the random-restart PCs I see turn out to be running too hot.

When a processor hits its thermal limit (usually around 100°C for Intel, 95°C for AMD), the motherboard shuts everything down to stop permanent damage. On some machines that means a clean shutdown. On most it means the PC just drops dead and reboots.

The classic pattern: the PC restarts after 20 to 30 minutes of use, or when you do something that pushes the processor hard (gaming, video editing, a Zoom call with 15 people). It works fine for a while after the restart because the shutdown gave it a chance to cool down. Then it happens again.

Open up the case if it's a desktop and look at the CPU heatsink. If it's clogged with dust and pet hair, that's your answer. Blow it out with a can of compressed air. For laptops, dust builds up inside the vents over time and there's no easy way to clear it without opening the bottom panel. I wrote about this in more detail in my post on how to stop your laptop overheating.

Desktops are easier to fix yourself. Laptops, if you're not comfortable opening them up, are worth bringing in. A good clean and fresh thermal paste usually sorts it.

Windows Update is doing it

Before you start worrying about hardware, check this first. Windows 10 and 11 will restart your PC to finish installing updates, and if you've been putting them off, it'll eventually pick a moment and do it anyway. Usually the worst possible moment.

If the PC restarted once and it's been fine since, go to Settings > Windows Update and check the update history. If you see something installed around the time of the restart, that was almost certainly it. Annoying, but not a fault.

The August 2026 update has been causing a few restart-related issues on certain hardware. If this started happening after the 12th of August, have a look at that post.

You can change when Windows is allowed to restart by going to Settings > Windows Update > Advanced options and setting your active hours. It won't restart during those hours. Doesn't fix the problem, but at least it won't interrupt you mid-afternoon.

The power supply is failing

This is the second most common cause on desktops, and the hardest for most people to diagnose at home because a failing PSU doesn't always give you obvious signs.

A power supply that's on the way out can still power the PC most of the time. But when the system draws more current than the PSU can provide (a game starts, a render kicks off, sometimes even just a few Chrome tabs), it can't keep up and the PC loses power. It looks like a restart, but it's actually a momentary power cut.

Signs the PSU might be the problem:

• Restarts happen under load, not when the PC is idle
• The PC occasionally shuts off completely rather than restarting (no blue screen, just off)
• You notice a faint burning or hot plastic smell from the back of the case
• The PC is more than four or five years old and has been running hard

Cheap PSUs that came bundled with budget cases are the worst offenders. They're rated for a certain wattage but can't actually sustain it under real conditions, especially once they age. Replacing the PSU is one of the cheaper desktop repairs. A decent 550W unit from Corsair or EVGA costs about £50-60, and fitting it takes half an hour.

A driver is crashing

Windows has a setting called "automatically restart" that's on by default. When the system hits a critical error (a proper crash), it restarts instead of sitting on a blue screen. That means you might be getting blue screens and never actually seeing them because the PC reboots too quickly.

The graphics driver is usually the one to blame. Nvidia, AMD, and Intel all release drivers with bugs. A driver that worked fine yesterday can start crashing after a Windows update changes something underneath it.

To check, right-click the Start button and open Event Viewer. Go to Windows Logs > System and look for entries marked "Critical" or "Error" around the time of the restart. You're looking for things like "The system has rebooted without cleanly shutting down first" (Event ID 41, Kernel-Power). If you see "Display driver stopped responding and has recovered" (Event ID 4101), the graphics driver is crashing.

If it is the graphics driver, update it from the manufacturer's website (Intel DSA, Nvidia GeForce Experience, AMD Adrenalin). If you recently updated the driver and that's when the crashes started, roll it back: Device Manager > Display adapters > right-click your graphics card > Properties > Driver > Roll Back Driver.

Your RAM might be going bad

Faulty RAM causes some of the most confusing symptoms. The PC might restart randomly, freeze, blue screen with different stop codes each time, or just behave strangely in ways that don't point to any one thing.

RAM doesn't usually fail completely all at once. A single bit goes wrong in one of the modules, and whenever the system tries to use that specific part of memory, it crashes. Because different programs use different parts of RAM at different times, the crashes seem random. One day it's fine, next day it restarts three times.

Windows has a built-in memory test. Search for "Windows Memory Diagnostic" in the Start menu and run it. It'll restart the PC and test the RAM before Windows loads. If it finds errors, you'll see a notification after the next boot. The test is fairly basic though. If you want a thorough check, download Memtest86 from memtest86.com, write it to a USB stick, and boot from it. Leave it running for a few hours. Any errors at all means at least one stick needs replacing.

If you've got two sticks of RAM, you can try running with one at a time to work out which is faulty. Swap them between slots too, in case the slot itself is the problem.

Your storage is dying

A failing hard drive or SSD can cause restarts when Windows tries to read or write to a damaged area of the disk. This is more common with traditional spinning hard drives than SSDs, but SSDs do fail too.

Warning signs alongside the restarts: the PC takes longer to boot than it used to, programs hang for no reason, or you see "delayed write failed" errors. If you hear clicking or grinding from a desktop hard drive, that drive is in trouble.

Check the drive's health by opening a Command Prompt as administrator and typing wmic diskdrive get status. If it says anything other than "OK" for every drive, you've got a problem. For a more detailed picture, a free tool called CrystalDiskInfo reads the drive's SMART data and gives you a green/yellow/red status. Yellow or red means it's time to replace the drive and get your data off it while you still can. My hard drive failure guide covers what to do next.

Malware (less common than you'd think)

People often assume random restarts mean a virus. It's possible but it's not the most likely cause. Most modern malware doesn't want your PC to restart, because a restarting PC isn't mining cryptocurrency or sending spam for the attacker.

That said, rootkits and some types of ransomware can cause instability that leads to restarts. If the restarts started after you downloaded something dodgy, installed software from a site you don't recognise, or clicked a link in a suspicious email, run a scan. Windows Defender does a decent job, but for a second opinion, download Malwarebytes (the free version is fine for a one-off scan) and run a full system scan. If it finds something, let it quarantine the files and see if the restarts stop.

How to narrow it down

If you're trying to work out which of the above is causing your restarts, start with these:

1. Check Windows Update history to rule out a forced update restart
2. Check temperatures with HWMonitor while using the PC normally. If the CPU is hitting 90°+ before a restart, it's heat
3. Open Event Viewer and look at Critical/Error entries around the restart times. The event log usually tells you exactly what happened
4. Run Windows Memory Diagnostic to check for RAM faults
5. Run CrystalDiskInfo to check drive health

Between those five checks you can usually narrow it to one cause. If Event Viewer shows Kernel-Power errors with no preceding driver crash or overheating, the PSU is the most likely culprit, and that's harder to test at home without spare parts.

When to get it looked at

If you've been through the above and can't pin it down, or if the restarts are getting worse, it's worth getting someone to look at it. PSU faults in particular need testing with a multimeter or a known-good replacement unit. Intermittent RAM faults sometimes don't show up in short tests. And if you're seeing different blue screen codes each time, that can mean a motherboard issue, which takes proper diagnosis to confirm.

Leaving it too long is the main risk. A PC that's overheating can damage the CPU or GPU permanently. A PSU that's failing can take other components with it when it goes. A dying hard drive will lose your data if it gives out completely. Better to catch it early.

If you're in St Helens or nearby and your PC is restarting on its own, I can usually work out what's causing it within the first half hour. PC repairs in St Helens, evenings and weekends.

Mark

Mark has been fixing computers since the late '90s and went self-employed in 2008. Based in St Helens since 2013, he works evenings and weekends from his home in Laffak — friendly, affordable repairs for PCs, laptops, and Macs. See reviews on Google