A PC that powers on, gets partway through booting, then restarts itself β sometimes over and over in a loop β is a different and often more revealing fault than one that simply won't turn on at all. Because it's getting somewhere before failing, the point at which it restarts tells you a great deal about the likely cause. This guide breaks down the most common reasons and how to work through them systematically.
Understanding Where the Restart Happens
The single most useful diagnostic question is: at what point does it restart?
- Restarts immediately, before any display output β points to a power delivery, motherboard or CPU issue
- Restarts during POST (manufacturer logo screen) β often RAM, GPU, or an overclock/BIOS setting issue
- Restarts during the Windows loading spinner, before login β often a corrupted boot file, driver issue, or failing storage drive
- Restarts after reaching the desktop, under load or randomly β often overheating, a failing PSU under load, or a driver crash
Pay close attention to this pattern before trying fixes, as it narrows the likely cause considerably.
Restarts Before or During POST
Check for an Automatic Restart on Failure Setting
Some BIOS versions are configured to automatically reboot after a POST failure rather than halting with an error message, which can make a real hardware fault look like an unexplained restart loop.
- Enter BIOS/UEFI setup (Del or F2 at startup).
- Look for a setting often called "Reboot on POST failure," "Automatically restart" or similar, usually under Advanced or Boot settings, and disable it temporarily.
- This lets you actually see any error message or beep code instead of it flashing past in a loop.
Check RAM and GPU Seating
Poorly seated RAM or a graphics card is a very common cause of restart loops at this stage.
- Power off, unplug, and earth yourself.
- Reseat RAM sticks firmly, one at a time if you have more than one, testing with a single stick first.
- Reseat the graphics card, ensuring it clicks fully into the PCIe slot and all power connectors are firm.
Check for a Recent Overclock or XMP Change
If the restarts started right after enabling XMP/EXPO or manually overclocking, this is very likely the cause.
- Enter BIOS/UEFI and reset to default settings (often an option called "Load Optimised Defaults" or similar), or physically reset CMOS if you can't get in reliably enough to navigate the menu.
- Boot normally, then reintroduce XMP/EXPO one step at a time rather than an aggressive manual overclock, to find a stable configuration.
Warning: An unstable overclock or overly aggressive XMP profile doesn't just cause restarts β sustained instability can occasionally corrupt data being written at the time of a crash. If you suspect an overclock is the cause, resolve it before relying on the PC for anything important.
Restarts During Windows Loading
Try Safe Mode
- Interrupt the boot process three times in a row by holding the power button during startup (as soon as the Windows logo appears, hold power for a few seconds to force off, then repeat) β this triggers Windows' Automatic Repair environment.
- Choose Troubleshoot > Advanced options > Startup Settings > Restart, then select Safe Mode.
- If the PC boots and stays stable in Safe Mode (which loads minimal drivers), this strongly points to a driver or software conflict rather than a hardware fault.
Run Startup Repair and Check Boot Files
- From the same Advanced options menu, choose Startup Repair and let Windows attempt to fix boot-related issues automatically.
- If that doesn't resolve it, boot from a Windows installation USB, choose Repair your computer > Troubleshoot > Advanced options > Command Prompt, and run:
-
bootrec /fixmbr-bootrec /fixboot-bootrec /rebuildbcd - These rebuild the boot configuration data and master boot record, fixing many "stuck in a loop before login" scenarios caused by corrupted boot files.
Check the Drive for Failure
A failing hard drive or SSD can cause exactly this symptom β Windows starts loading, hits a read error on a critical system file, and restarts.
- From the same Command Prompt environment, run
chkdsk C: /f /rto scan for and attempt to repair drive errors (this can take a long time on a large or heavily damaged drive β let it run). - If possible, connect the drive to another working PC via a USB adapter and check its SMART health status using CrystalDiskInfo β a "Caution" or "Bad" status strongly suggests the drive itself is failing and needs replacing.
Restarts After Reaching the Desktop
Check Temperatures
- Boot into Safe Mode or briefly into Windows if it's stable enough, and open HWiNFO to monitor CPU and GPU temperatures.
- If temperatures spike close to 100Β°C under even light load, this points to a failed or poorly seated CPU cooler, dried-out thermal paste, or a case with inadequate airflow β see our guides on reapplying thermal paste and preventing overheating for detailed fixes.
Check the PSU
Restarts specifically under load (gaming, heavy CPU tasks) but stability at idle strongly suggest a power supply struggling to deliver under demand.
- Check Windows Reliability Monitor (search "Reliability" in the Start menu) for a pattern of unexpected shutdowns correlating with high-load activity.
- Test with a different, known-good PSU if available β this is the most conclusive way to confirm or rule out a PSU fault.
- See our dedicated guide on testing and replacing a failing PSU for detailed steps.
When to seek professional help: If restarts happen unpredictably with no clear pattern β not tied to load, not tied to a specific boot stage, and unresolved after checking RAM, GPU seating, temperatures and the PSU β this can indicate a failing motherboard, a marginal power connector fault, or in rare cases a failing CPU. These are harder to diagnose without swap-in test components, and a professional repair service will typically have known-good spare parts to isolate the fault far faster than continued guesswork.
Was this guide helpful? Explore more Power & Boot Issues guides, or browse all Desktop PC Repair articles.