Windows BSOD Troubleshooting: A Complete Stop Code Guide

Diagnose blue screen errors, unexpected restarts, and Windows stop codes methodically—using the crash evidence, recent changes, supported recovery tools, and targeted hardware tests.

What a blue screen actually means

A blue screen of death—also called a bug check, stop error, kernel error, or BSOD—appears when Windows detects a condition from which it cannot safely continue. Windows records a stop code and restarts or shuts down to reduce the risk of further corruption or data loss. The root cause can be a defective or incompatible driver, failing hardware, firmware, security software, corrupted system files, unstable overclocking, or another low-level fault.

The stop code’s symbolic name, such as IRQL_NOT_LESS_OR_EQUAL, describes the type of failure Windows detected. It does not automatically identify the guilty component. A driver named on the screen or in a dump can be the cause, a victim of corrupted memory, or simply the component active when another failure surfaced.

What to do immediately after a BSOD

1. Photograph the screen

   Capture the complete stop code, QR code, percentage, and any line beginning with “What failed.” Note the time and what the computer was doing—starting, waking, gaming, copying files, installing an update, or sitting idle.

2. Let Windows finish collecting information

   If progress is moving, allow the device to restart normally. Do not repeatedly cut power unless the screen is genuinely frozen for an extended period or the hardware is overheating.

3. Disconnect only newly added nonessential hardware

   After shutdown, remove a recently connected dock, USB device, expansion card, or peripheral if the crash began immediately after it was added. Keep the keyboard, mouse, display, and essential storage connected.

4. Decide whether it was isolated or recurring

   One crash followed by normal operation should be documented and monitored. Repeated crashes, boot loops, changing stop codes, file corruption, or failures during light activity justify immediate diagnosis.

Build an evidence timeline

Use the crash pattern to narrow the cause

Safe BSOD repair path

1. Return the system to known-stable settings

Undo the single most relevant recent change. Remove unsupported overclocking, undervolting, memory profiles, and experimental BIOS settings. If you changed several things, return to manufacturer defaults and reintroduce changes only after the system is stable.

2. Install supported Windows updates

Open Settings → Windows Update, install applicable quality updates, and restart. If the BSOD started immediately after a specific update, review Microsoft’s known issues and use the supported uninstall or recovery option rather than disabling updates indefinitely.

3. Update, roll back, or reinstall the relevant driver

Target the device connected to the evidence. Use Windows Update first, then the exact PC model’s official support page or the hardware maker’s official site. If a previously stable device began crashing after an update, Device Manager → device Properties → Driver → Roll Back Driver is often safer than layering more packages onto the system.

4. Repair Windows components and protected files

If multiple Windows components fail or system corruption is suspected, open Terminal as administrator and run DISM before SFC:

Restart when repairs are reported. These commands can restore Windows components; they do not diagnose defective RAM, an unstable CPU, failing storage, or a bad third-party driver.

5. Test memory without guessing

Search for Windows Memory Diagnostic, choose to restart and check, then review the result in Event Viewer under the System log for MemoryDiagnostics-Results. A reported error is significant. A passing short test does not rule out intermittent RAM, memory-controller, motherboard, heat, or timing problems.

If memory was recently installed, confirm the modules and configuration are supported by the system or motherboard. Power off and follow the manufacturer’s instructions before reseating hardware. Test at default settings before blaming Windows.

6. Check storage, free space, and physical connections

Maintain adequate free space for Windows, paging, updates, and crash dumps. Review the storage manufacturer’s health tool and firmware guidance. On serviceable desktops, power down before checking data and power cables. Do not rely on a single “healthy” summary if the system also logs disk, controller, NTFS, or reset errors.

The online scan checks the file system; it does not prove the physical drive is healthy. Back up before any repair that writes heavily to a suspected failing disk.

7. Use a clean boot to isolate third-party conflicts

A clean boot starts Windows with a reduced set of non-Microsoft services and startup programs. If stability returns, re-enable items in groups to identify the conflicting security product, filter driver, hardware utility, overlay, VPN, tuning tool, or updater. Restore normal startup when testing is complete.

When Windows will not start: use WinRE

Windows Recovery Environment can open automatically after repeated startup failures. You can also hold Shift while choosing Restart, use Settings → System → Recovery → Advanced startup, boot from a recovery drive, or select Repair my PC from official Windows installation media.

1. Choose Troubleshoot → Advanced options.
2. Use Startup Settings → Restart → Safe Mode when a recent driver or service is suspected.
3. Use Uninstall Updates when the failure began with a recent Windows update.
4. Use System Restore to roll system files and settings back without targeting personal documents.
5. Use Startup Repair when Windows cannot complete the startup process.
6. Consider Reset or reinstall only after backups and less destructive options are exhausted.

Common stop-code families

Always compare the symbolic name and numeric bug-check code with Microsoft’s Bug Check Code Reference. The four parameters attached to a bug check can change its meaning and are primarily intended for technical analysis.

Crash-dump analysis with WinDbg

For recurring crashes, an IT professional or advanced user can open the dump in Microsoft WinDbg and run:

Read the bug-check code, parameters, stack, involved modules, and failure bucket together. Do not uninstall or replace a driver solely because its filename appears in one automated summary. Compare multiple dumps: repeated evidence pointing to the same third-party module is more persuasive than a different blamed module in every crash.

When to stop and seek hardware or professional help

• The device crashes during startup, Safe Mode, or official diagnostics.
• Memory diagnostics report errors or files repeatedly become corrupt.
• Storage health warnings, controller resets, unusual noises, or disappearing drives appear.
• Temperatures, fans, power delivery, liquid damage, or physical damage are involved.
• Different stop codes continue at manufacturer-default settings.
• The dump implicates storage, encryption, endpoint security, or another business-critical kernel component.
• You cannot access backups, the BitLocker recovery key, or a reliable recovery environment.

Reduce future BSOD risk

• Keep Windows, firmware, applications, and evidence-relevant drivers supported and current.
• Download drivers and firmware only from Windows Update or official manufacturers.
• Back up important data and keep recovery keys accessible.
• Create a restore point before high-risk driver or system changes.
• Change one low-level component at a time and record versions.
• Use stable manufacturer-supported BIOS and memory settings.
• Maintain cooling, free disk space, surge protection, and healthy storage.
• Preserve crash photos and dumps until the problem is resolved.

BSOD decision table

Frequently asked questions

References and further reading

• Microsoft Support: Resolving Blue Screen errors in Windows
• Microsoft Support: Windows Recovery Environment
• Microsoft Support: Recovery options in Windows
• Microsoft Learn: Bug Check Code Reference
• Microsoft Learn: Analyze bug check stop-code data
• Microsoft Support: WHEA_UNCORRECTABLE_ERROR