STATUS_REGISTRY_HIVE_RECOVERED 0x8000002A fix

When you see this error

You boot Windows 10 or 11 and get a blue screen or a notification in Event Viewer saying STATUS_REGISTRY_HIVE_RECOVERED (0x8000002A). The full message reads: "The registry hive (file): %hs was corrupted and it has been recovered." The %hs placeholder is the actual hive file path — often something like \SystemRoot\System32\Config\SAM or \??\C:\Users\\ntuser.dat.

This error pops up after an unexpected shutdown, a disk write failure, or a corrupt driver write to the registry. It's not a fatal error — Windows already recovered the hive from a backup copy — but it's a serious warning that the file system or disk might be failing.

Root cause

The registry hive files (SYSTEM, SOFTWARE, SAM, SECURITY, DEFAULT, ntuser.dat) are memory-mapped files. When Windows writes to them, it does so in 4K blocks. If the power cuts out mid-write, or if the disk has bad sectors, a block gets partially written — that's corruption. On next boot, the Session Manager (SMSS) detects the checksum mismatch and rolls back to the last known good copy stored in %SystemRoot%\System32\config\RegBack (pre-Windows 10 1803) or uses a transactional log replay (Windows 10 1803+).

What's actually happening here is that the recovery succeeded, but the underlying cause—disk errors, failing RAM, or a rogue driver writing garbage into the hive—still hasn't been fixed. You're seeing the symptom, not the problem.

What to do

Check which hive was recovered. Open Event Viewer (eventvwr.msc), go to Windows Logs > System, and filter by event ID 15 (source: Kernel-General). The log entry shows the exact hive path. Write it down.
Run a disk check on the system drive. Open a Command Prompt as Administrator and run:
chkdsk c: /f /r
You'll need to restart. This scans for bad sectors and file system errors. The /r flag locates bad sectors and recovers readable data. Skip the /r if you're in a hurry — but don't; fix it properly.
Check the hive itself for lingering corruption. If the recovered hive is a user hive (ntuser.dat), create a new local admin account and migrate data. For system hives, run sfc /scannow and DISM /Online /Cleanup-Image /RestoreHealth to repair system files. These don't fix registry corruption directly, but they fix the framework around it.
Test RAM with Windows Memory Diagnostic. Type mdsched.exe in Run, choose Restart now and check for problems. Bad RAM causes registry writes to land in wrong cells, producing corruption that looks like a disk issue. If you get errors, swap the faulty stick.
Restore from a backup if chkdsk finds nothing. If the error persists but your disk and RAM are clean, restore the hive from your last system backup. If you use File History or a full image backup, mount it and copy the hive files from %SystemRoot%\System32\config in the backup to the same location in Windows Recovery Environment (WinRE).

If it still fails

If the error keeps coming back after chkdsk and RAM tests pass, the disk itself is dying. Check SMART status with a tool like CrystalDiskInfo. If you see Reallocated Sector Count or Current Pending Sector in yellow or red, replace the drive immediately. Also check the motherboard's SATA or NVMe controller drivers — outdated or buggy drivers can corrupt registry writes. Update them from the manufacturer's site, not from Windows Update.

One more thing: if you're on Windows 10 build 1803 or later, the automatic backup copies in RegBack are missing by design (Microsoft stopped making them). In that case, the "recovery" was a transactional log replay, and if the log itself was corrupt, you got a stale or empty hive. That's when you need a known-good backup from before the corruption happened.