Random restarts usually come from overheating, a failing power supply, corrupted Windows files, or a driver that crashes the system

When your PC restarts without warning, it is almost never a single cause—it is one of four or five common problems, and you can narrow it down by looking at the pattern. Does it happen during heavy use, or at idle? Does Windows show an error before it goes dark, or does it just cut to black? Does it restart when ready, or does it sit for a few seconds first? The answers tell you whether to check your cooling system, your power supply, your drivers, or your Windows installation.

The restart happens because Windows detects a critical error and shuts down to prevent hardware damage. Your job is to find out what triggered that error—and the good news is that most of the time, the answer is something you can fix yourself without opening the case or calling a technician.

Key Takeaways

  • Overheating is the most common cause of random restarts; check that your CPU and GPU fans spin freely and that dust is not blocking your vents.
  • A failing power supply often causes restarts during heavy use (gaming, video editing) because it cannot deliver enough power under load.
  • Corrupted Windows files or a driver update can trigger a restart loop; Windows Safe Mode lets you test whether a recent driver is the culprit.
  • Check the Event Viewer log (search "Event Viewer" in Windows) to see what Windows recorded in the seconds before each restart.
  • If restarts happen only during one specific task, the problem is usually that task's driver or the hardware it uses, not your whole system.

How to check if your PC is overheating

Overheating is the easiest problem to spot and often the easiest to fix. Dust buildup on your CPU cooler or case fans starves them of airflow, and the system restarts when the CPU or GPU hits a temperature threshold—usually 90 to 100 degrees Celsius, depending on your hardware.

read a free temperature monitor like HWiNFO or Core Temp and run it while you work. Watch the CPU and GPU temperatures. If either one climbs above 80 degrees Celsius during normal use, or above 85 degrees during gaming or video work, overheating is likely your culprit. If the temperature spikes right before a restart, that confirms it.

To fix it, shut down your PC, unplug it, and open the case (or remove the side panel if it is a prebuilt). Look at your CPU cooler and case fans. If they are caked with dust, use a can of compressed air to blow the dust out—spray in short bursts and let the fans spin freely. Do not touch the fans while they are spinning. Clean the intake vents on the front and bottom of the case too. Reassemble, plug in, and run your temperature monitor again.

If the temperatures are still high after cleaning, your cooler may have dried out thermal paste (the gray compound between the CPU and the cooler). Replacing it is a 15-minute job if you are comfortable opening the case, or a $30 to $50 job at a repair shop.

Testing whether your power supply is failing

A power supply that is wearing out often delivers enough power for everyday tasks but fails under load. You will notice restarts during gaming, video editing, or any task that makes your CPU and GPU work hard at the same time. The restart happens when ready with no warning—the system just goes dark.

The first test is straightforward: does the restart happen only during heavy use, or does it happen at random times when you are just browsing? If it is only during heavy use, your power supply is the prime suspect. If it happens at idle too, look at the other causes first.

A power supply is not something to repair yourself—it contains high-voltage components that can injure you even when unplugged. If you think yours is failing, you have two options. You can buy a replacement (they cost $50 to $150 depending on wattage) and install it yourself if you are comfortable opening the case and unplugging cables. Or you can take the PC to a repair shop and have them swap it out. Either way, note your current power supply's wattage (it is printed on the label on the back of the unit) so you buy one with the same or higher wattage.

How to check the Event Viewer log for error codes

Windows records what happened in the seconds before each restart. The Event Viewer is a built-in tool that shows you those logs, and it often points directly to the problem.

Search for Event Viewer in the Windows search box and open it. On the left side, click Windows Logs, then click System. You will see a long list of events. Look for entries marked Error or Critical that happened at the time of your restart. Click on one and read the details. Look for a line that says Source—that tells you what part of Windows or what driver caused the problem.

Common sources include Kernel-Power (usually overheating or power supply), nvlddmkm (NVIDIA graphics driver), amdkmdag (AMD graphics driver), or the name of a specific device driver. If you see the same source before every restart, that is your problem. Write down the source name and move to the next section.

Fixing driver crashes and Windows file corruption

If the Event Viewer shows a graphics driver or another device driver as the source, the driver is crashing and taking Windows with it. The fix depends on when the crashes started.

If the restarts began right after you updated Windows or installed a new driver, boot into Windows Safe Mode and uninstall the recent update. To enter Safe Mode, restart your PC and press F8 or Shift+F8 repeatedly as it boots (the timing varies by PC—if F8 does not work, try holding Shift while clicking Restart in the Power menu). Choose Safe Mode with Networking. Once you are in Safe Mode, open Settings > System > About > Advanced system settings > Device Manager. Find the device that matches the source from Event Viewer (for example, your graphics card under Display adapters), right-click it, and choose Uninstall device. Restart normally. Windows will reinstall a generic driver, and you can test whether the restarts stop.

If the restarts continue in Safe Mode, the problem is Windows file corruption, not a driver. You can run a repair using the command line. Open Command Prompt as Administrator (search for it, right-click, and choose Run as administrator) and type sfc /scannow. Let it run—it will take 10 to 15 minutes. If it finds corrupted files, it will repair them automatically. Restart and test.

When to replace hardware versus when to call a technician

If cleaning your cooler and fans stops the restarts, you are done. If the Event Viewer points to a driver and uninstalling it fixes the problem, you can reinstall a fresh version from the manufacturer's website. Both of those are safe DIY fixes.

If the problem is your power supply, you can replace it yourself if you are comfortable unplugging cables and screwing in a new unit—it is straightforward, but if you are unsure, a repair shop will do it for $50 to $100 in labor. If the Event Viewer shows Kernel-Power errors and cleaning your cooler did not help, or if the sfc /scannow command found corrupted files it could not repair, take the PC to a technician. Those problems require either hardware replacement or a Windows reinstall, and a technician can do both faster and with less risk of data loss.

Frequently Asked Questions

Can a virus cause random restarts?

Viruses can cause restarts, but it is rare. If you suspect malware, run a full scan with Windows Defender (built into Windows) or a free tool like Malwarebytes. If the scan finds nothing and the Event Viewer shows a driver or hardware source, the restarts are almost certainly hardware-related, not malware.

What if my PC restarts only when I play a specific game?

That usually means the graphics driver is crashing under the load that game creates. Try lowering the graphics settings in the game (resolution, texture quality, frame rate cap) to see if it stops. If it does, your GPU is either overheating or your power supply cannot handle the load. Check temperatures first; if they are normal, your power supply is likely the problem.

Does a restart loop mean I need a new PC?

No. A restart loop (where Windows tries to boot, fails, and restarts again) usually means corrupted Windows files or a driver that crashes during startup. Boot into Safe Mode, uninstall recent updates or drivers, and run sfc /scannow. If that does not work, a technician can reinstall Windows without losing your files.

Should I disable automatic restart to see the error message?

Yes, if you want to read the error before the restart. Go to Settings > System > About > Advanced system settings > Startup and Recovery, uncheck Automatically restart, and click OK. The next time an error occurs, Windows will show you the message instead of restarting. Write down what it says and search for it online, or bring it to a technician.

Can I prevent restarts by updating Windows?

Sometimes. If your restarts began after a Windows update, the update itself may be the problem—uninstall it in Safe Mode as described above. If your Windows installation is several months old, check for new updates in Settings > Update & Security > Windows Update. New updates often fix driver conflicts and stability issues. Install them and restart normally.