Fix 0XC00002C5: STATUS_DATA_ERROR on Windows

You're staring at a blue screen that says EXCEPTION (0XC00002C5) or STATUS_DATA_ERROR, and the machine either reboots or sits there until you power it off. This isn't one of those errors you can click past. It means the kernel received data it couldn't trust, and it stopped the whole system rather than risk writing garbage to disk.

The stop code is 0xC00002C5 which maps to STATUS_DATA_ERROR. Microsoft's own documentation classifies it as a message integrity failure, which is why your first instinct (Googling "Windows 10 reinstall") is the wrong one. A clean Windows install won't fix bad RAM.

Cause 1: Faulty or mismatched RAM (this is the one, nine times out of ten)

I've seen this error come back to bad memory more than every other cause combined. A single failing DIMM in a 4-slot board can throw STATUS_DATA_ERROR once a week or once an hour, depending on which bank is flaky. The classic trigger: you upgraded RAM six months ago, and the new sticks were a different speed or brand than the originals. Mixed kits sometimes train at the wrong timings and produce silent corruption.

Test your RAM properly. Don't trust the Windows Memory Diagnostic alone; it passes marginal sticks that MemTest86 fails.

1. Download MemTest86 from memtest86.com and write it to a USB stick. The free version is fine.
2. Boot from the USB. On modern UEFI machines you may need to disable Secure Boot temporarily.
3. Let it run through a full pass. A full pass on 16GB takes about 45 minutes on a typical desktop.
4. If you see any red errors, stop. One error = bad stick. MemTest86 doesn't produce false positives.

If you get errors, pull all but one stick and test each in slot A2 (the second slot from the CPU on most boards — check your manual, this matters). When you find the bad one, replace the whole kit, not just the failed stick. Mixed kits cause this error too.

Also check RAM speed in BIOS. If your board defaults to 2133 MHz and your kit is rated 3200, enable XMP or DOCP. If XMP is on and you're getting 0xC00002C5, turn it off as a test. Unstable overclocks produce exactly this signature.

Cause 2: A driver writing bad data to the kernel

After RAM, the next most common source is a driver that's handing the kernel malformed structures. Antivirus filter drivers, VPN tunnel adapters, and cheap USB 3.0 controllers top the list. The error usually shows up during a specific action — plugging in a device, connecting a VPN, running a specific app.

Check the minidump first. On a default Windows 10 or 11 install, dumps live here:

C:\Windows\Minidump

Open the newest .dmp with BlueScreenView (free, tiny, no install). Look at the module listed at the top of the stack. If it's ntoskrnl.exe only, that tells you nothing. If you see nvlddmkm.sys, rtwlane.sys, iaStorAC.sys, or a filter driver from your AV, that's your culprit.

Fix it:

1. Note the driver name from BlueScreenView.
2. Identify which piece of hardware or software owns it. rtwlane.sys is Realtek Wi-Fi. iaStorAC.sys is Intel RST storage. Filter drivers usually belong to third-party security tools.
3. Get the newest driver from the manufacturer's site, not Windows Update. Windows Update lags by months on storage and network drivers.
4. If it's a security product, uninstall it fully — most have a removal tool. Rebooting into Safe Mode first helps because filter drivers resist uninstall while running.

If you can't pin it down, enable Driver Verifier. Fair warning: this can make the machine unbootable until you disable it. Do it from Safe Mode and note the command to disable.

verifier /standard /all
verifier /reset

The first command enables verification on all non-Microsoft drivers. Reboot, reproduce the crash, then check the new minidump — Verifier forces the offending driver to bugcheck instead of silently corrupting memory. Run the second command from Safe Mode to turn it off.

Cause 3: Failing storage or a corrupted system file

Less common than RAM or drivers, but this is where you end up if the first two checks pass. A drive with reallocated sectors can return bad blocks to the filesystem, and Windows treats that as corrupt data. If your SSD is older than four years or your SMART data shows reallocated sectors climbing, this is likely.

Check SMART first. CrystalDiskInfo is the easiest tool — green is fine, yellow means back up now, red means order a replacement today.

Then run these in an elevated Command Prompt, in this order:

chkdsk C: /scan
sfc /scannow
DISM /Online /Cleanup-Image /RestoreHealth

chkdsk /scan is the non-destructive version. It reports problems without locking the volume. If it finds errors, schedule a full check with chkdsk C: /f /r and reboot — it'll run before Windows loads.

sfc /scannow should finish with "Windows Resource Protection did not find any integrity violations." If it says it fixed corrupt files, run it a second time. If it can't fix them, DISM /RestoreHealth pulls clean copies from Microsoft's servers. Run it before sfc if sfc fails twice in a row.

One more thing to check

If you're on Windows 11 22H2 or later and this error appears right after a feature update, check for a bad update. Roll it back:

wusa /uninstall /kb:KB_NUMBER

You can find the pending or recent KB in Settings > Windows Update > Update history. Feature updates have shipped with kernel-level bugs before — 22H2 had a known issue with certain Intel storage drivers that produced STATUS_DATA_ERROR during sleep transitions.

Quick reference: what to try, in order

Work top to bottom. Don't skip the RAM test because it takes an hour — it's the fix 60% of the time, and you'll waste a whole day chasing drivers on a machine that just needs a new stick. If you've tested RAM, cleared the driver, and verified the disk, and the error still comes back, capture a fresh minidump and post it to a forum with the hardware list. That's the point where a second set of eyes on the stack trace is worth the wait.