First move: write down the stop code
The text on the blue screen matters. A stop code such as DRIVER_IRQL_NOT_LESS_OR_EQUAL or MEMORY_MANAGEMENT, and especially any filename ending in .sys, narrows the cause from “something is wrong with my PC” to a specific subsystem. If the screen disappears too fast, you can retrieve the same information afterwards from Event Viewer or from the minidump file. Do not start reinstalling Windows before you have read the code.
What a blue screen actually is
Windows runs most code in user mode, where a crash kills one program and nothing else. Drivers and core system components run in kernel mode, where there are no guard rails. When something in kernel mode does something impossible — writes to memory it does not own, waits at the wrong interrupt level, receives corrupt data from hardware — Windows cannot safely continue, because carrying on risks writing corrupted data to disk.
So it stops. Deliberately. The blue screen is a controlled shutdown with a diagnostic report, not a malfunction in itself. That reframing is useful, because it tells you that the underlying fault is nearly always in a driver or in hardware, not in Windows.
Reading the evidence
Find the stop code after the fact
If the machine rebooted before you could read the screen, open Event Viewer (search for it in the Start menu), go to Windows Logs then System, and look for a critical event from the source “Kernel-Power” or an error from “BugCheck”. The BugCheck entry contains the stop code and parameters.
Alternatively, go to Settings then System then About then Advanced system settings, click Settings under Startup and Recovery, and untick “Automatically restart”. The next blue screen will then stay on screen until you power the machine off, giving you time to read and photograph it.
Match the stop code to a category
Codes are not arbitrary. DRIVER_IRQL_NOT_LESS_OR_EQUAL, SYSTEM_THREAD_EXCEPTION_NOT_HANDLED and KMODE_EXCEPTION_NOT_HANDLED point at drivers, and usually name the offending .sys file. MEMORY_MANAGEMENT, PAGE_FAULT_IN_NONPAGED_AREA and IRQL_NOT_LESS_OR_EQUAL often indicate RAM problems. CRITICAL_PROCESS_DIED and UNEXPECTED_STORE_EXCEPTION frequently mean storage or filesystem corruption. WHEA_UNCORRECTABLE_ERROR is a hardware error reported by the CPU itself and usually means genuine hardware trouble, often related to overclocking or heat.
This categorisation does not give you the answer, but it tells you which of the following steps to do first.
Test the memory
RAM faults cause a wide and confusing range of blue screens because corrupted memory can break anything. Run Windows Memory Diagnostic — press Windows + R, type mdsched.exe, and let it restart and test.
Windows’ own test is quick but not thorough. If it passes and you still suspect memory, MemTest86 on a USB stick run for several full passes is the real standard, and it catches faults the built-in tool misses. If you have multiple RAM sticks, testing one at a time identifies which one is bad.
Check the disk
Open an administrator command prompt and run chkdsk /f /r on your system drive; it will schedule itself for the next restart. Then run sfc /scannow to verify system files, followed by DISM /Online /Cleanup-Image /RestoreHealth if sfc reports problems it cannot fix.
For an SSD, also check its health using the manufacturer’s utility — Samsung Magician, Crucial Storage Executive, or a general tool such as CrystalDiskInfo. An SSD approaching end of life produces intermittent crashes well before it fails outright.
Roll back or update the suspect driver
If the blue screen named a .sys file, search for that filename to identify which driver it belongs to. Common culprits are graphics drivers (nvlddmkm.sys for NVIDIA, atikmdag.sys for AMD), network drivers, and third-party utilities such as RGB lighting or fan control software, which install kernel drivers of highly variable quality.
If crashes started after an update, roll the driver back in Device Manager. If the driver is old, update it from the hardware manufacturer’s site rather than Windows Update. For graphics drivers specifically, a clean installation using the vendor’s own removal option resolves problems that a simple update does not.
Boot into Safe Mode to isolate software
Hold Shift while clicking Restart, then choose Troubleshoot, Advanced options, Startup Settings, Restart, and pick Safe Mode with Networking. Safe Mode loads a minimal set of drivers.
If the machine is stable in Safe Mode, the fault is in software or a third-party driver, not in core hardware. You can then use msconfig to disable non-Microsoft services in batches and find the culprit by elimination. If it blue-screens even in Safe Mode, suspect hardware.
Check temperatures and undo any overclock
Install HWMonitor or a similar tool and watch CPU and GPU temperatures under load. Sustained CPU temperatures above roughly 95°C, or a GPU above about 85°C, cause instability that presents as random crashes.
If you have overclocked anything — CPU, GPU, or RAM via an XMP or EXPO memory profile — return it to stock settings and test. Memory profiles are the most commonly overlooked overclock: enabling XMP is running the RAM beyond its default specification, and not every combination of board, CPU and kit is stable at the advertised speed.
If the blue screens started at a clearly identifiable moment, that is your strongest clue. New hardware, a Windows feature update, a driver update, or a newly installed program are all worth reversing before doing any deeper diagnosis. System Restore to a point before the change takes ten minutes and often ends the investigation.
Reading a minidump properly
Windows writes a small crash dump to C:\Windows\Minidump each time. These are readable with WinDbg from the Microsoft Store, but the friendlier option for most people is a small free utility called BlueScreenView, which lists every crash with its stop code and highlights the driver files implicated.
What you are looking for is a pattern. One crash naming a random system file tells you little. Five crashes over two weeks all naming the same third-party driver tells you exactly what to remove.
If there are no dump files at all, check that crash dumps are enabled: Advanced system settings, Startup and Recovery, and confirm “Small memory dump” or better is selected. Also confirm the page file is not disabled, since dumps are written through it.
Avoid “driver updater” and “PC optimiser” programs, particularly when troubleshooting crashes. They install drivers that do not match your hardware, which is an excellent way to convert an occasional blue screen into a machine that will not boot. Get drivers from the hardware manufacturer or from Windows Update only.
When it is genuinely hardware
Some patterns point strongly at failing hardware rather than software: crashes that happen in Safe Mode, crashes during the Windows installer, WHEA errors, crashes that correlate with heat or load, and crashes that continue after a clean Windows installation.
In those cases the usual order of suspicion is RAM first (easiest to test and to swap), then the power supply (particularly in an older machine or one with a recently upgraded graphics card), then storage, then motherboard. A failing power supply is underdiagnosed because it produces symptoms that look like everything else — random reboots, crashes under load, and instability that comes and goes with ambient temperature.
Is a single blue screen something to worry about?
Not really. An isolated crash after a driver update, a power interruption, or an unusual load is common and not a sign of a failing machine. What matters is repetition. If it happens once and never again, note the stop code and move on. If it happens weekly, investigate properly.
Will reinstalling Windows fix it?
It fixes software causes — corrupted system files, bad drivers, conflicting software — and it fixes nothing at all if the cause is hardware. Because a reinstall costs hours, it is worth doing the memory test and the driver investigation first. If crashes continue after a clean install, you have at least proven the problem is physical.
What does the percentage counter on the blue screen mean?
It is the progress of writing the crash dump to disk. Letting it reach 100% before the machine restarts is what gives you a minidump to analyse, so it is worth waiting rather than forcing a power-off.
Can overheating cause blue screens?
Yes. Modern CPUs throttle and then shut down rather than crash, but GPUs, memory and voltage regulators are less protected, and sustained heat causes errors that surface as crashes. Check temperatures under load, clean out dust, and confirm fans are actually spinning. On a laptop, elevated temperatures with no obvious dust often mean dried-out thermal paste.
Why do crashes only happen during games?
Because games load the GPU, CPU, memory and power supply far harder than anything else, so they expose marginal hardware first. The usual suspects are graphics driver problems, GPU overheating, insufficient or failing power supply, and unstable memory overclocks. Test by running the machine at stock clocks with a monitored temperature log.
The bottom line
Read the stop code, use it to pick a starting category, then test memory, check the disk, and investigate the named driver. Safe Mode separates software causes from hardware ones quickly, and minidump analysis turns a vague problem into a named file. Reinstalling Windows is a reasonable step, but it belongs near the end of the process rather than the beginning.
