Blue Screen on Your PC? How to Find What Caused It, Step by Step

In this guide (11 sections)

• What exactly happens?
• 1. Write down the stop code
• 2. Check the list of past crashes in Reliability Monitor
• 3. Find the driver file named in the crash dump
• 4. Undo recent changes and overclocks
• 5. Test the RAM
• 6. Reinstall the graphics driver cleanly
• 7. Repair Windows files and check the disk
• 8. Update the BIOS and chipset drivers
• 9. Check temperatures and the power supply
• Real cases: what other users ran into

A blue screenBlue screen (BSOD): The full-screen error Windows shows just before it restarts after a serious crash. On recent Windows 11 updates it is black instead of blue. More in our Windows & performance glossary →Close ✕ (often called the Blue Screen of Death, or BSOD) means Windows hit an error so serious that it stopped everything and restarted, rather than risk damaging your files. It’s a protection, not a sign that the PC is broken. A single blue screen after an update or a new driver is often nothing. Repeated ones always have a cause, and Windows leaves clues that point to it. This guide shows you how to read those clues, then how to try the fixes that solve most cases, from the quickest to the most involved.

What exactly happens?

What you see	Most likely cause	Where to look
The PC restarted by itself, and you didn’t manage to read the screen.	You need the stop code first	Steps 1 and 2
The blue screens started after a new app, driver or Windows update.	That change	Steps 3 and 4
The blue screens started after turning on XMP or EXPO, or changing BIOS settings.	Memory or processor settings	Step 4
Each blue screen shows a different stop code.	Usually RAM, or settings that make the PC unstable	Steps 4 and 5
The screen says WHEA_UNCORRECTABLE_ERROR.	Hardware instability, mostly the processor	WHEA_UNCORRECTABLE_ERROR guide, then step 4
The screen says VIDEO_TDR_FAILURE, or names a file like nvlddmkm.sys or amdkmdag.sys.	The graphics driver or graphics card	VIDEO_TDR_FAILURE guide, then step 6
The screen says DPC_WATCHDOG_VIOLATION.	A driver that took too long, often storage or a background app	DPC_WATCHDOG_VIOLATION guide, then step 7
The PC crashes only after a long time gaming, and the PC is hot or loud.	Overheating or the power supply	Step 9

1. Write down the stop code

Every blue screen shows a stop code: a name in capital letters with underscores. Many also show a “What failed” line with a file name. These two clues decide where to look next.

Where to look on the blue screen

Older Windows shows it in blue, with a sad face. Recent Windows 11 updates show a simpler black screen, with the same two clues near the bottom.

• 1The stop code
  The name in capital letters, here VIDEO_TDR_FAILURE. It tells you what kind of error caused the crash. Write it down or take a photo.
• 2"What failed" (not always shown)
  A file name. It often points to the driver involved: here nvlddmkm.sys, the NVIDIA graphics driver.

How to do it: next time it happens, take a photo of the screen with your phone. If the PC restarts too quickly to read it, you can stop it from restarting by itself:

1. Press the Windows key, type advanced system settings and open it.
2. Under Startup and Recovery, click Settings.
3. Untick Automatically restart and click OK.

The next blue screen will stay on screen until you restart the PC yourself. You can tick the box again once you’ve found the cause.

2. Check the list of past crashes in Reliability Monitor

Windows keeps a day-by-day list of every crash. It’s the quickest way to see the stop codes you missed and to spot a pattern.

How to do it: press the Windows key, type reliability and open View reliability history (Reliability MonitorReliability Monitor: A built-in Windows page that lists every crash and problem day by day, with the stop code. More in our Windows & performance glossary →Close ✕). Each red X marks a day when the PC crashed. Click it, then look for “Shut down unexpectedly” or “Windows was not properly shut down” and click View technical details. The stop code appears as a short number after “Code”: for example 116 is VIDEO_TDR_FAILURE, 124 is WHEA_UNCORRECTABLE_ERROR and 133 is DPC_WATCHDOG_VIOLATION. Searching “stop code” plus the number finds the name of any other code.

What to look for:

• Always the same stop code: follow the guide for that code (the table above links them).
• A different stop code each time: suspect the RAM, its settings or an overclock first (steps 4 and 5).
• Crashes that began on a certain day: check what you installed or changed that day.

3. Find the driver file named in the crash dump

At each blue screen, Windows tries to save a small report called a crash dumpCrash dump (minidump): A small file Windows saves at each blue screen, recording what was running when it crashed. More in our Windows & performance glossary →Close ✕ in the folder C:\Windows\Minidump. A free tool can read it and show which driver file was running when the PC crashed.

How to do it:

1. Install WhoCrashed (free for home use) or BlueScreenView (free).
2. Open it. It reads the crash dumps by itself and lists the stop code and the driver file for each crash.
3. Look up the file name. Common ones are below.

File name	What it belongs to
nvlddmkm.sys	NVIDIA graphics driver
amdkmdag.sys or atikmpag.sys	AMD graphics driver
igdkmd64.sys	Intel integrated graphics driver
storahci.sys, stornvme.sys, iaStorAC.sys	Drivers for disks and SSDs
ntoskrnl.exe	The core of Windows itself. It’s rarely the real cause: it often points to RAM, settings or hardware.

A name you don’t recognize often belongs to an app: RGB lighting, audio enhancers, VPNs, anti-cheat or antivirus programs install their own drivers. Search the file name to find which app it is, then update or uninstall that app as a test.

If the Minidump folder is empty, Windows didn’t manage to save the report. That happens when the PC switches off instead of showing a blue screen, which usually points to the power supply or to overheating (step 9).

4. Undo recent changes and overclocks

Most blue screens that appear “out of nowhere” start right after something changed. Undoing it is the quickest test.

How to do it:

• A new driver: open Device Manager, double-click the device, go to Driver and click Roll Back Driver, or install the previous version from the maker’s website.
• A new app: uninstall it for a few days, especially RGB lighting, fan control, audio “enhancer”, VPN or game anti-cheat software.
• A Windows update: go to Settings → Windows Update → Update history → Uninstall updates and remove the latest one as a test.
• BIOS settings: load the BIOS defaults (Load Optimized Defaults; the keyboard key to press is shown at the bottom of the BIOS screen). This also turns off XMP or EXPO and any automatic overclockingOverclocking: Running a part faster than its standard speed. More speed, but crashes if the part can't handle it. More in our Windows & performance glossary →Close ✕, such as MCE on Intel motherboards or PBO and Curve Optimizer on AMD. Our guide on how to clear CMOS shows another way to reset the BIOS if you can’t get into it.
• A graphics card overclock: reset MSI Afterburner or your graphics driver’s tuning page to default.

If the crashes stop with everything at default, turn things back on one at a time. XMP or EXPO is often the one that needs attention: our guide on a PC that won’t start after enabling XMP explains how to make it stable, usually by choosing a slightly lower memory speed.

5. Test the RAM

Faulty or unstable RAM is one of the most common causes of repeated blue screens, especially when every crash shows a different stop code.

How to do it:

1. Quick test: press the Windows key, type Windows Memory Diagnostic, open it and choose Restart now and check for problems. The result appears after Windows starts again.
2. Thorough test: MemTest86Memory test (MemTest86): A test that writes and reads the RAM over and over to catch errors that cause random crashes. More in our Windows & performance glossary →Close ✕ (free) starts from a USB stick and tests the RAM outside Windows. Let it run at least 4 full passes. Any error at all is a problem.
3. If there are errors: test with XMP or EXPO off, then test each module on its own in the slot your manual recommends. A module that fails on its own is faulty; if each passes alone but the pair fails, the speed setting is too high for the pair.

Our guide on which RAM slots to use shows where each module should go, and RAM not running at full speed explains the speed settings.

6. Reinstall the graphics driver cleanly

If the stop code mentions VIDEO, or the crash dump shows the file nvlddmkm.sys, amdkmdag.sys or atikmpag.sys, the graphics driver is involved. A normal update installs over the old driver; a clean install removes everything first.

How to do it:

1. Download the latest driver for your graphics card from NVIDIA, AMD or Intel, plus the free tool Display Driver Uninstaller (DDUClean driver install (DDU): Removing every trace of the graphics driver with a free tool, then installing a fresh copy. More in our Windows & performance glossary →Close ✕).
2. Unplug the network cable or turn off Wi-Fi, so Windows doesn’t install its own driver in the meantime.
3. Start Windows in Safe Mode: hold Shift while clicking Restart, then go to Troubleshoot → Advanced options → Startup Settings → Restart and press 4.
4. Run DDU, choose your graphics card brand and click Clean and restart.
5. Install the driver you downloaded, then reconnect the internet.

If the newest driver crashes and the previous one didn’t, stay on the previous one until the next release. Our guide on VIDEO_TDR_FAILURE covers the rest of the graphics card checks.

7. Repair Windows files and check the disk

Damaged Windows files, an SSD with old firmware (its built-in software) or a failing disk can all cause blue screens, often with stop codes about storage or with no clear driver named.

How to do it:

1. Repair Windows files (SFC and DISMSystem file check (SFC and DISM): Built-in Windows commands that find and repair damaged Windows files. More in our Windows & performance glossary →Close ✕): right-click the Start button, open Terminal (Admin) and run DISM /Online /Cleanup-Image /RestoreHealth, then sfc /scannow. Restart when they finish.
2. Check the disk’s health: the free tool CrystalDiskInfo shows a health status for each disk. Anything other than “Good” is a warning sign.
3. Update the SSD’s firmware: each SSD maker has a free app for it (for example Samsung Magician, WD Dashboard, Crucial Storage Executive). Back up your files first.
4. Keep free space: leave at least 10–15% of the Windows disk free.

8. Update the BIOS and chipset drivers

A newer BIOS often fixes stability problems, especially on newer processors and with fast RAM. Chipset drivers tell Windows how to work with the motherboard.

How to do it: find your exact motherboard model (type System Information in the Start menu and look at “BaseBoard Product”). On the maker’s support page for that model, compare the newest BIOS with yours (“BIOS Version/Date” in the same window). Follow the maker’s instructions exactly and don’t switch off the PC during the update. A BIOS update also resets the BIOS settings, so you may need to turn XMP or EXPO on again afterwards. If you’re not comfortable doing it, a repair shop can do it quickly. Install the chipset driver from the same page, or from AMD or Intel.

9. Check temperatures and the power supply

Blue screens that only happen after a long time gaming can come from heat, and sudden switch-offs without any blue screen usually come from the power supply.

How to do it: watch your temperatures while you play (how to monitor temps while gaming). If the PC switches off with no blue screen at all, our guide on a PC that turns off while gaming covers power and heat step by step, and our guide to the signs of a failing power supply shows how to test it.

If you’ve tried every step and the blue screens continue, a clean reinstall of Windows tells you whether software is to blame: if the crashes continue on a fresh Windows with only the essential drivers, the cause is hardware.

Real cases: what other users ran into

These are real threads from PC forums, summarized in our own words. Each one links to the original.

Real caseCore i9-10900K · ROG Strix Z490-F · 4 × 8 GB DDR4 at 4000 MHz (XMP) · RTX 3090SolvedBlue screens gone after removing two of the four RAM modulesRead what happened

In December 2021, a few weeks after moving to Windows 11, the PC kept crashing with IRQL_NOT_LESS_OR_EQUAL. A forum helper noticed that the RAM was running a 4000 MHz XMP profile, well above the speed the processor officially supports. The person who posted reset all overclocks to default and turned XMP off. MemTest86 then kept failing with all four modules installed, but passed with two. With two modules, the blue screens stopped, and in January 2022 the thread was marked solved.

Takeaway: a fast XMP profile on four modules is a lot to ask of the processor’s memory controller. See step 5.

Source: Windows 11 Forum

Real caseRyzen 7 7700X · ASUS X670E-A · 32 GB DDR5 · RTX 3080SolvedAn audio app's driver behind repeated crashesRead what happened

In August 2023, a new PC had game crashes and IRQL_NOT_LESS_OR_EQUAL blue screens. Three different memory tests found no errors, and removing the RAM overclock didn’t help. Following the forum helpers’ advice, the person who posted ran Driver Verifier, a Windows tool that puts third-party drivers under stress, and the PC crashed at once, naming several driver files. One belonged to Boom3D, an audio equalizer app. After it was removed, the blue screens stopped. A month later the person who posted confirmed there had been no more crashes.

Takeaway: the driver software of an everyday app can crash Windows. See step 3.

Source: Windows 11 Forum

Real caseAMD PC with DDR5 (EXPO) · WD SN770 system SSDSolvedRepeated blue screens after a clean Windows install, gone after an SSD updateRead what happened

In October 2025, a freshly installed Windows 11 kept crashing with CRITICAL_PROCESS_DIED, sometimes just switching off. Changing the RAM settings didn’t help, and disk tests on all seven drives plus eight passes of a memory test found nothing. The person who posted then updated the firmware of the WD SN770 SSD that held Windows, and a pending Windows update installed around the same time. After three days with no crashes, they reported the problem fixed. It isn’t certain which of the two updates did it.

Takeaway: update the SSD’s firmware and Windows before replacing parts. See step 7.

Source: Windows 11 Forum

Frequently asked questions

Can a blue screen damage my PC?

The blue screen itself doesn’t. It’s Windows stopping on purpose to avoid damage. Repeated crashes are worth fixing, though: a crash in the middle of saving a file or installing an update can leave that file damaged.

I only had one blue screen. Should I worry?

Usually not. A single crash after a driver or Windows update is common. Note the stop code in Reliability Monitor and only start troubleshooting if it happens again.

Why is my blue screen black?

Since the second half of 2025, Windows 11 shows a black version of the same screen. It means exactly the same thing and still shows the stop code, so everything in this guide applies.

The PC restarts or switches off, but I never see a blue screen. Is it the same problem?

Not always. If Reliability Monitor shows a stop code, it was a blue screen that flashed past. If it only says the PC shut down unexpectedly, with no stop code, a power supply problem or overheating is more likely. See our guide on a PC that turns off while gaming.

Will reinstalling Windows fix it?

Only if the cause is software. It’s a useful last test: if a fresh Windows with only the essential drivers still crashes, the cause is hardware, and you’ve saved yourself further guesswork.

Can a virus cause blue screens?

It’s possible but rare today. Third-party antivirus and anti-cheat drivers cause far more blue screens than viruses do. A scan with Microsoft Defender’s offline scan is a reasonable check if nothing else explains the crashes.