Error Messages

Kernel Data Inpage Error (0x7A): Causes and Fixes

Updated October 2, 2026 •7 min read

Your PC freezes, then drops to a blue screen with the stop code KERNEL_DATA_INPAGE_ERROR. It might happen once, or every time you boot. A kernel data inpage error has a narrow, specific meaning, and more often than most blue screens, it points at the drive Windows is running from. That makes the order of your next steps matter.

What the kernel data inpage error means

Windows doesn’t keep everything in RAM. Parts of memory that aren’t in active use get written out to the paging file (pagefile.sys) on your system drive. When Windows needs that data again, it reads the page back in. That’s an “in-page” operation.

Stop code 0x7A (KERNEL_DATA_INPAGE_ERROR) means Windows tried to read a page of kernel data back from disk and the read failed. Because the kernel can’t keep running without that data, it halts the system.

So the core question is: why couldn’t Windows read from the disk?

Common causes

Bad sectors or a failing drive. This is the cause to rule out first. If the area holding part of the paging file (or a system file) has degraded, reads fail intermittently, and the crash comes back.
A loose or faulty connection. A worn SATA cable, a loose SATA power connector, or an NVMe SSD that isn’t fully seated in its M.2 slot can drop reads under load.
Faulty RAM. Bad memory can corrupt data on its way in or out of the paging file.
Paging-file or file-system corruption, or outdated storage drivers or SSD firmware. Less common, and often a symptom of one of the above.

Find the error status (optional, but useful)

After a crash, Windows logs the stop code with four parameters. Open Event Viewer (right-click Start > Event Viewer) > Windows Logs > System and look for a BugCheck entry (Event ID 1001) around the time of the crash. For 0x7A, the second parameter is usually the I/O status code, and a few values are worth knowing:

0xC000009C or 0xC000016A: the disk reported bad blocks.
0xC000009D: the device wasn’t connected, which points to cabling or the controller not seeing the disk.
0xC0000185: an I/O device error, often cabling or connection related.

While you’re in the System log, filter for the source disk. Entries such as Event ID 7 (“has a bad block”), Event ID 51 (“an error was detected on device during a paging operation”), or Event ID 153 (an I/O operation was retried) all point to the storage path, not RAM or software.

Step 1: Check drive health before anything else

Many guides tell you to run chkdsk /r right away. On a drive with bad sectors, that’s the wrong first move. /r reads every sector on the disk and keeps retrying weak areas, which is hours of stress on hardware that may already be failing.

Instead, run a SMART check first with CrystalDiskInfo or a similar tool:

Good status with zero reallocated, pending, and uncorrectable sectors → the drive is probably not the cause. Move on to connections and RAM.
Caution or Bad, or nonzero Current Pending Sector Count or Reallocated Sector Count → treat the drive as failing. Go straight to Step 2.
On NVMe SSDs, look at Media and Data Integrity Errors and Available Spare. Rising errors or low spare are both bad signs.

If Windows crashes too often to run a check, connect the drive to another computer as a secondary drive and check it there.

Step 2: Back up before you repair

If SMART shows any warning, or you don’t have a current backup, copy your important files to a different drive now, before running any repair tools. Start with what’s irreplaceable: documents, photos, project files. If the drive is clicking, freezing, or dropping out during the copy, stop and read hard drive failure warning signs before you continue.

Step 3: Reseat connections

Power off and unplug the computer.

Desktop with a SATA drive: reseat the SATA data cable at both ends and try a different cable and motherboard port. Check that the power connector is firmly seated.
NVMe SSD: remove it, check the slot for dust, reseat it firmly, and tighten the retaining screw.
Laptop: if the drive is user-accessible, reseat it.

If the crashes stop after this, the connection was the problem.

Step 4: Test memory

Press Win + R, type mdsched.exe, and choose Restart now and check for problems. The Windows Memory Diagnostic runs before Windows loads and reports the results after you log back in. For a more thorough test, a bootable tool like MemTest86 runs multiple passes. Any memory errors at all mean a RAM stick needs to be replaced.

Step 5: Repair system files and the paging file (healthy drive only)

Once SMART is clean and your data is backed up, run these from an administrator Command Prompt:

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

chkdsk C: /f will ask to schedule the check at the next restart. Type Y and reboot. Only move on to chkdsk C: /r if the drive is healthy and backed up.

To rebuild the paging file: open Settings > System > About > Advanced system settings. Under Performance, click Settings > Advanced > Virtual memory: Change. Select No paging file, click Set, and restart. Then go back and choose System managed size and restart again. Windows creates a fresh pagefile.sys.

Also check your PC or SSD maker’s support site for storage driver, SSD firmware, and BIOS updates.

When to stop and get help

Stop troubleshooting and power the computer down if:

SMART shows pending or uncorrectable sectors, or the count keeps rising.
The drive clicks, grinds, or beeps, or disappears from the BIOS.
The crash happens every boot and you can’t get your files off.
Event Viewer is full of disk bad-block errors.

Every boot and crash makes a failing drive retry the same weak areas, and each attempt can make recovery harder.

The bottom line

A kernel data inpage error means Windows couldn’t read part of its own memory back from disk. Before you reach for repair commands, check the drive’s SMART health, because bad sectors and loose connections cause it far more often than software does. If the drive is healthy, reseat cables, test RAM, and run sfc and chkdsk /f. If it’s failing, back up first and skip the full-surface scans.

If the crashes have locked you out of a drive with files you can’t lose, our hard drive recovery service images failing drives on specialized hardware for one flat $1,200 price for a single device. Shipping is free and prepaid, and you pay nothing unless we recover your data.