PC Randomly Shuts Off or Restarts? How to Find the Real Cause

A computer that shuts off or restarts on its own, with no error message and no warning, is one of the more unsettling problems because there's usually nothing on screen to search for — it just goes dark or reboots and comes back like nothing happened. The lack of an error message doesn't mean there's no evidence, though. Windows logs these events, and the pattern of when they happen narrows the cause considerably before you touch any hardware.

Shutdown vs. restart — this distinction matters

These two symptoms often get lumped together, but they point toward different categories of cause.

• A full shutdown (power goes off completely, fans stop, you have to press the power button again) more often points toward a power delivery problem — the PSU, a power connector, or a hard thermal cutoff.
• An automatic restart (system reboots itself and comes back to Windows) is frequently a driver-level crash configured to restart rather than show a blue screen, or Windows' automatic reboot setting hiding a BSOD from view entirely.
• Knowing which one you're actually experiencing changes where you start looking.

Check if a BSOD is happening and getting hidden

Windows can be configured to automatically restart after a blue screen so fast that you never actually see it, which makes a driver crash look identical to a random restart.

• Go to Settings > System > About > Advanced system settings > Startup and Recovery and uncheck "Automatically restart."
• With this off, an actual crash will show the blue screen and stay there instead of rebooting immediately, giving you the stop code to work with.
• Also check Event Viewer > Windows Logs > System for a Kernel-Power Event ID 41 — this specifically logs that the system restarted without a clean shutdown, which confirms an unexpected power loss or hard crash occurred.

Rule out overheating as a hard cutoff

Modern CPUs and GPUs have a thermal protection feature that shuts the system down instantly if temperatures reach a dangerous threshold, which happens with essentially zero warning from the user's perspective.

• Install HWMonitor or HWiNFO and check the maximum recorded temperatures logged before a shutdown, not just the current reading.
• If CPU or GPU temps were sitting near or at their thermal limit right before the shutdown, heat is very likely the direct cause rather than a coincidence.
• This pattern is especially telling if shutdowns only happen under load — gaming, rendering, or other demanding tasks — rather than at idle.

Test the power supply's actual capacity

A power supply that's underpowered for the system, or one that's aging and no longer delivering its rated wattage reliably, causes shutdowns specifically when demand spikes.

• Shutdowns that happen at the exact moment of a demand spike — launching a game, a GPU ramping up, multiple drives spinning up simultaneously — point strongly toward the PSU being unable to supply enough power at that instant.
• Add up your components' power draw (CPU, GPU, drives) and compare against the PSU's rated wattage — a system running close to or over the PSU's capacity has little headroom for the power spikes that happen under real-world load.
• PSUs also degrade with age; a unit that was adequate when new can fail to deliver its full rated output several years later, especially cheaper units with lower-quality internal components.

Check for a loose or failing power connection

Sometimes the cause isn't capacity but a physical connection issue that interrupts power briefly, long enough to trigger a shutdown.

• Reseat the 24-pin motherboard connector and the CPU power connector (usually 4 or 8-pin near the top of the motherboard) firmly on both ends.
• Check PCIe power connectors going to the GPU if applicable — a partially seated connector here can cause shutdowns specifically under GPU load.
• Inspect cables for any visible damage, especially near the connector where repeated flexing over time can cause an internal wire to fray without being visible from the outside.

Faulty RAM can trigger shutdowns, not just BSODs

While RAM errors more commonly produce blue screens, severe memory faults can sometimes cause a hard shutdown or restart instead, particularly if the error occurs in a way the system can't recover from cleanly.

• Run Windows Memory Diagnostic with the extended test option, or MemTest86 overnight for a more thorough check.
• If multiple RAM sticks are installed, testing with one stick at a time in different slots can isolate a single bad module or a bad slot rather than assuming all memory is suspect.

Surge protectors and outlets — the overlooked culprit

Occasionally the problem isn't inside the computer at all.

• A surge protector nearing the end of its life can cause intermittent power interruptions that look exactly like a PC hardware fault.
• Plug the PC directly into the wall, bypassing any surge protector or power strip, and see if the shutdowns stop — this single test isolates the computer from the electrical path feeding it.
• If shutdowns happen alongside other devices flickering or resetting on the same circuit, the building's electrical supply itself may be the actual source.

When it points to the motherboard

If the PSU tests adequate, connections are solid, RAM is clean, and temperatures are reasonable, but shutdowns persist, a motherboard-level fault becomes the more likely remaining cause — a failing capacitor, a VRM issue, or a short somewhere that only manifests under specific conditions. This is less common than the causes above, but it's the point where component-level bench testing becomes more useful than continuing to test in the original system.

Dealing with random shutdowns or restarts that seem to have no pattern? Bring it in for a full diagnostic — we'll check power delivery, thermals, and memory to find the actual cause instead of guessing.