Games Keep Crashing on PC? Sort the Crash by What It Looks Like First (2026)

By LK Wood IV · 2026-07-11 · ~9 min read · St. Louis County, MO

Search “games keep crashing pc” and you get the same twenty-item checklist over and over: update drivers, verify files, disable XMP, check temps, run SFC. It’s not wrong - but handing you all twenty at once, in no order, is why people “try everything” and still crash. The list ignores the one thing that actually narrows it down: what the crash looks like.

A game that quietly closes to the desktop, a blue screen with an error code, and a whole PC that hard-locks or reboots mid-game are three different failures with three different culprits. Sort your crash into the right one first and you can skip three-quarters of that checklist. Here’s the fork, then how to read the actual crash instead of guessing at it.

Step 1: what does the crash actually look like?

Before you change one setting, put your crash in a row. This is the fork nobody leads with, and it’s the whole game:

The two most useful splits inside that table: one game versus every game, and crash-to-desktop versus the whole machine going down. If it’s one game closing to desktop, your hardware is almost certainly fine and you’re wasting time in the BIOS. If it’s every game plus blue screens or a dead machine, the problem is system-wide - memory, heat, or power - and no amount of reinstalling one game will fix it.

Step 2: read the crash - don’t guess at it

Windows writes down every crash. Almost nobody looks, which is how the twenty-item checklist survives. Two free, built-in tools turn a guess into a lead.

Reliability Monitor is the friendly one. Press Start, type Reliability History, and open it. You get a calendar with a red X on every crash. Click the day, and each failure shows a plain-language line plus a View technical details link. It’s the fastest way to see whether your crashes cluster right after a Windows update, a driver install, or a new program - which often is the answer.

Event Viewer is the precise one. Press Start, type Event Viewer, then open Windows Logs then Application, and find the red Application Error entry (Event ID 1000) stamped at your crash time. Two fields matter:

• Faulting Application Name - which .exe actually died. If it’s the game, the crash is in the game or its renderer. If it’s an overlay or helper app, you just found your culprit.
• Faulting Module Name - the specific file that broke. A graphics DLL like nvwgf2umx.dll (NVIDIA), atidxx64.dll (AMD), or d3d12.dll points at the GPU driver or DirectX. A third-party module - RGB software, a capture tool, antivirus - points at that program.

One more clue: the Exception Code. 0xc0000005 is an access violation, and when games throw it repeatedly it usually means bad memory or an unstable overclock - which tells you to go straight to the XMP test below. (Event ID 1001 next to 1000 is just Windows’ crash-reporting entry for the same event.)

If the game keeps its own crash logs - many do, in the game’s folder or Documents - read those too; they often name the exact subsystem that failed.

The game-level fixes (one game, or crash-to-desktop)

If a single game closes to desktop, or every game crashes to desktop without a blue screen, work this list. These are free and none of them touch your hardware:

• Update the GPU driver - properly. Get it from NVIDIA, AMD, or Intel directly, not Windows Update (which often ships an older build). If crashes persist, do a clean install: wipe the current driver with Display Driver Uninstaller (DDU) in Safe Mode, then install the latest fresh. Corrupt or half-updated drivers are the single most common cause of crash-to-desktop.
• Verify the game’s files. On Steam: right-click the game then Properties then Installed Files then Verify integrity of game files (other launchers have a “repair” button). This re-downloads anything corrupt - the top cause of one game crashing while the rest are fine.
• Kill the overlays. Discord, the NVIDIA App/GeForce Experience, AMD Adrenalin, Steam overlay, Xbox Game Bar, and MSI Afterburner / RivaTuner (RTSS) all inject code into the game’s render pipeline. Under DirectX 12 especially, a single overlay can corrupt the command queue the game sends to the GPU and crash it - even if that overlay works fine in other games. Turn them all off, confirm the crash is gone, then re-enable one at a time to find the offender.
• Delete the shader cache and update the game. A bad patch or a stale shader cache causes crashes that a driver update alone won’t fix. Clear the cache (NVIDIA Control Panel or the game’s own option) and make sure the game is current.
• Check anti-cheat and compatibility. Some anti-cheat systems clash with overlays, virtual machines, or Windows’ Core Isolation. For older games, try running as administrator or in a compatibility mode.

The system-level fixes (every game, BSOD, or the whole PC goes down)

If every game crashes, you get blue screens, or the machine hard-locks or reboots, the problem is your hardware or its settings. Run these in order - cheapest and highest-hit first:

1. Disable XMP / EXPO and test. This is the highest-hit fix on the system side. Your RAM’s XMP (Intel) or EXPO (AMD) profile is a factory overclock; it can pass everyday use and fail under a game’s memory load. In the BIOS, set memory to default/Auto (JEDEC) speed and test. Crashes stop? Your kit couldn’t hold that overclock - re-enable XMP one speed tier lower, or nudge DRAM voltage up slightly.
2. Test the RAM itself. If it still crashes at default speed, run MemTest86 from a USB stick for several passes (or Windows Memory Diagnostic). Any error means a bad module - reseat the sticks, try one at a time, and RMA if a stick fails.
3. Rule out heat. A CPU or GPU hitting its thermal limit shuts down to protect itself, which reads as a mid-game crash or hard-lock. Watch temperatures with HWiNFO or HWMonitor during play; if the CPU or GPU is pinned at 90C+ when it dies, that’s your answer - see CPU overheating and GPU overheating.
4. Suspect the power supply. An aging or undersized PSU sags under the millisecond power spikes a modern GPU pulls when a game ramps up, dropping the whole system - a hard reboot with no blue screen is the classic signature. Our PSU failure symptoms guide covers how to confirm it and why the “paperclip test” won’t.
5. Remove every overclock and undervolt. GPU core/memory overclocks and CPU undervolts that seemed stable in a benchmark routinely crash under real game loads. Set everything back to stock and test before you blame a part - and if the blue screen you keep getting is CLOCK_WATCHDOG_TIMEOUT, that revert is not just a diagnostic step, it is the fix.

The quick-win order

If you’d rather have a sequence than a diagnosis, this catches most crashes fast:

1. Read the crash in Reliability Monitor / Event Viewer - note the faulting module and whether it’s one game or all.
2. One game only? Verify its files, update the GPU driver, disable overlays. Stop here - it’s not your hardware.
3. Every game or BSODs? Disable XMP/EXPO and test. This alone fixes a large share of “random” crashes.
4. Still crashing? Watch temps and test RAM with MemTest86.
5. Hard reboots with no blue screen? Look hard at the power supply.

Most crashes end in the first three steps. The people who “tried everything” almost always skipped step 1 - they never read the crash, so they were changing settings blind. Sixty seconds in Event Viewer usually tells you which half of this guide you actually needed.

If your game isn’t crashing but stuttering - running but hitching, with uneven frame times rather than a clean shut-down - that’s a different problem; start with why your PC stutters in games.