How to Fix the NTFS FILE SYSTEM Error in Windows 10

What the stop code actually says

The message names the driver that died — ntfs.sys, the kernel component that reads and writes NTFS volumes — not necessarily the culprit. When Windows cannot trust what the disk returns, it stops with this code rather than write garbage, which makes 0x00000024 a protective stop, a close cousin of KERNEL DATA INPAGE ERROR: something low in the storage stack failed, and the system refused to continue blind.

In practice three causes cover the overwhelming majority of cases: corruption of NTFS metadata after a power cut or hard shutdown; a buggy or outdated storage controller driver, sometimes triggered by third-party disk or virtual-drive software; and a drive that is physically dying, with bad sectors where the driver expects readable metadata. RAM is the rarity here — unlike MEMORY MANAGEMENT codes, this one rarely traces to faulty memory, so focus on the storage path first.

• File system corruption after a power loss, forced shutdown or crashed defrag
• Failing drive — bad sectors, reallocated sectors climbing in the SMART attributes
• Buggy or outdated SATA/NVMe controller or chipset driver
• Third-party disk utilities, old virtual-drive software or antivirus filter drivers hooking the stack
• A damaged system file that chkdsk and SFC will catch later in the sequence

The repair sequence

Start in Safe Mode if the blue screen loops before you can log in — hold Shift while clicking Restart, then Troubleshoot → Advanced options → Startup Settings. In a working system open an elevated Command Prompt and run chkdsk C: /f; answer Y to schedule the check at the next reboot, restart, and let it finish — interrupting chkdsk mid-run is how a repairable disk becomes a brick. If SMART shows reallocated sectors, or chkdsk reports bad clusters on every run, clone or back up the data first and plan a replacement drive instead of repairs.

Then repair the software side: sfc /scannow for broken system files, DISM /Online /Cleanup-Image /RestoreHealth if SFC itself reports corruption it cannot fix, and the latest chipset plus storage drivers from your motherboard or laptop vendor. Reboot and watch for a repeat. If the crash returns, pull the minidump — files in C:WindowsMinidump open in BlueScreenView or WinDbg and often name the exact third-party driver, which turns guesswork into a one-click uninstall. A full Microsoft Defender scan closes the malware branch, which is uncommon but cheap to rule out.

When to stop repairing

Two signals mean stop and salvage. First, chkdsk finding new bad sectors on every run, or SMART reporting Reallocated Sectors Count above zero and growing — the drive is dying; copy the data off while it still reads, because every repair pass spends the remaining life. Second, the same crash happening from a clean boot with updated drivers and a healthy drive — that points at deeper hardware: a failing SATA cable, an overheating controller or, after all, unstable RAM, and swapping the cable is the cheapest first test on a desktop.

Backups decide the cost of this error, not the repairs. If your documents, photos and project folders exist in a second copy, a blue screen is an hour of work; if they do not, the first move is imaging the drive before running anything on it. Windows 10 and 11 self-heal minor NTFS metadata errors, so a blue screen means the damage was beyond that — treat it as the disk waving, not as noise to mute.

Questions and Answers

Read next

All articles · FAQ