Few things are more alarming than your screen suddenly turning blue with a sad-face emoji and a wall of technical text before the computer restarts itself. A Blue Screen of Death (BSOD) is Windows' way of stopping immediately rather than risk data corruption when it hits a critical error, and while it looks catastrophic, the vast majority of causes are identifiable and fixable. This guide explains how to read the error, work through the most common causes systematically, and prevent it happening again.

Understanding what you're looking at

Modern Windows 10 and 11 BSODs display a stop code (such as IRQL_NOT_LESS_OR_EQUAL or MEMORY_MANAGEMENT) and sometimes a QR code alongside it. This stop code is the single most useful piece of information for diagnosis — write it down or photograph the screen before it restarts, as it disappears quickly.

  • A stop code naming a specific driver file (e.g. nvlddmkm.sys, ntoskrnl.exe) usually points directly at that driver or its associated hardware.
  • A generic code with no file named (e.g. KERNEL_SECURITY_CHECK_FAILURE) requires more investigation, often via the memory dump.
  • Random, varying stop codes each time suggest a hardware fault (commonly RAM or storage) rather than a single software driver.
  • The same stop code every time suggests a specific driver, application, or piece of hardware consistently triggering the same fault.

Reading the memory dump for more detail

  1. Open Settings > System > About, and note whether the machine is running 64-bit Windows.
  2. Download a free, reputable tool such as WhoCrashed or BlueScreenView, which reads the Windows minidump files stored at C:\Windows\Minidump and translates them into a readable summary of which driver or module was active when the crash occurred.
  3. Look at the "caused by driver" column — a repeated driver name here (rather than a generic Windows system file) is a strong lead.
  4. Cross-reference that driver name with recently installed hardware or software, particularly anything installed shortly before the crashes began.

Back up your important files as soon as possible after a BSOD, ideally to an external drive or cloud storage, before attempting fixes. Some causes of BSODs (particularly failing storage drives) can worsen suddenly, and you don't want to be troubleshooting while also racing to save irreplaceable files.

Step-by-step troubleshooting

1. Undo recent changes first

If the BSODs started after a specific event, that's your fastest lead:

  • Recently installed or updated a driver? Open Device Manager, find the device, right-click, choose Properties > Driver tab > Roll Back Driver (available for a limited time after an update).
  • Recently installed new hardware (RAM, a graphics card, an SSD)? Try removing it and testing whether the crashes stop, to isolate a compatibility or seating issue.
  • Recently installed a Windows Update? Go to Settings > Windows Update > Update history > Uninstall updates and remove the most recent quality update.
  • Recently installed new software? Uninstall it via Settings > Apps > Installed apps, particularly anything that installs low-level drivers such as antivirus suites, VPN clients, or virtualisation software.

2. Run Windows' built-in diagnostic tools

  1. Memory diagnostic: search "Windows Memory Diagnostic" in the Start menu and choose to restart and check for problems. This runs a RAM test on reboot and reports any errors found. For a more thorough test, download MemTest86 and run it from a bootable USB drive overnight, as Windows' built-in tool can miss intermittent faults.
  2. Check disk for errors: open Command Prompt as administrator and run chkdsk C: /f /r, then confirm the scan on next restart. This checks the drive for bad sectors and file system errors.
  3. System File Checker and DISM: run sfc /scannow followed by DISM /Online /Cleanup-Image /RestoreHealth in an administrator Command Prompt to repair corrupted system files — see our dedicated guide on repairing corrupted system files for the full process and what to do if SFC reports errors it can't fix.
  4. Driver Verifier (advanced): search "Driver Verifier" and enable it to stress-test third-party drivers, which can force a suspect driver to crash sooner and more informatively than waiting for it to happen naturally. Only use this if you're comfortable booting into Safe Mode afterwards to disable it again, as it can make a system less stable while active.

3. Update or roll back graphics drivers

Graphics drivers are one of the single most common BSOD causes, particularly nvlddmkm.sys (NVIDIA), atikmpag.sys (AMD), or igdkmd64.sys (Intel) related errors.

  1. Download the latest driver directly from NVIDIA, AMD, or Intel's official site rather than relying solely on Windows Update.
  2. Use the manufacturer's utility (e.g. NVIDIA app, AMD Software) to perform a "clean install" if available, which removes old driver remnants that can conflict with a fresh install.
  3. If crashes started right after a driver update, roll back instead via Device Manager as described above.

4. Check for overheating and power issues

  1. Download a monitoring tool such as HWMonitor or use the manufacturer's own utility to check CPU and GPU temperatures under load.
  2. Ensure vents aren't blocked, especially on laptops used on soft surfaces like beds or sofas.
  3. Clean dust from fans and heatsinks if the machine is more than a year or two old (see our dedicated dust and thermal maintenance guide).
  4. On desktops, check the power supply is adequate for the installed components, as an undersized or failing PSU can cause random BSODs under load.

5. Test RAM individually

If MemTest86 reported errors, or crashes are inconsistent:

  1. Power down and, following anti-static precautions, remove all but one RAM stick.
  2. Test the system with each stick individually in the same slot to identify a faulty module.
  3. Also try the same known-good stick in different slots to rule out a faulty motherboard slot rather than the RAM itself.

Common stop codes and what they usually mean

  • IRQL_NOT_LESS_OR_EQUAL: often a driver issue, occasionally faulty RAM
  • MEMORY_MANAGEMENT: usually points to RAM, though can be driver-related
  • CRITICAL_PROCESS_DIED: a core Windows process crashed, often after corrupted system files or a failed update
  • DPC_WATCHDOG_VIOLATION: frequently linked to storage drivers, particularly on SSDs with outdated firmware
  • KERNEL_SECURITY_CHECK_FAILURE: can indicate corrupted system files or, less commonly, failing hardware
  • WHEA_UNCORRECTABLE_ERROR: usually a hardware fault — CPU, RAM, or overheating are the most common causes

When to consider a clean reinstall

If you've worked through drivers, updates, memory, and disk checks without resolving the issue, and SFC/DISM haven't found repairable corruption, a clean reinstall of Windows via Settings > System > Recovery > Reset this PC (choosing to remove everything, after backing up) rules out any deep system corruption that's hard to pin down otherwise. This is a legitimate diagnostic step, not just a last resort — if BSODs stop after a clean install, you've confirmed the cause was software-side, and can reinstall drivers and applications one at a time to identify the culprit if it returns.

Still not fixed?

If BSODs continue even after a clean Windows reinstall, this strongly points to a hardware fault — most commonly RAM, storage, or occasionally the motherboard itself. At this point, professional diagnosis with dedicated hardware testing equipment is worthwhile, particularly if the laptop or desktop is still within warranty, since a manufacturer repair may be free where a DIY hardware swap would not be covered.

Was this guide helpful? Explore more Windows & macOS Tweaks guides, or browse all Operating System & Software articles.