BSOD-Related Data Loss

The Blue Screen of Death is what Windows displays when the operating system encounters a fatal error it cannot recover from. The screen itself isn’t the problem; it’s Windows shutting down to prevent further damage from continuing operation in a broken state. The BSOD includes a stop code, sometimes a QR code, and a brief description of what went wrong. The screen has been part of Windows since the early 1990s and has gone through several visual revisions, but its purpose has remained consistent: emergency stop when the OS detects something it cannot safely work around.¹

The blue screen on Windows 11 is now black

Microsoft changed the BSOD background color from blue to black starting with Windows 11 in 2021. The change was aesthetic, matching the rest of the Windows 11 design language. The functionality is unchanged: same stop codes, same QR codes, same emergency-shutdown behavior. The terminology in industry usage has stuck with “BSOD” regardless of whether the screen is actually blue or black. Some preview Windows builds have produced green screens during specific failures; these behave identically to the blue and black versions.

How BSODs cause data loss

BSODs cause data loss through several distinct mechanisms, and recognizing which one applies determines the right response:

• Lost work in memory: any unsaved data in open applications is lost when the BSOD forces a restart. A document being edited, a game in progress, a video being rendered. This is the most common but typically least severe form of BSOD data loss.
• Corrupted files mid-write: if a file was being written to disk when the crash occurred, the file may be left in a partially-written state. The result is file corruption on a single file rather than systemic damage.
• File system corruption: if the BSOD occurred during file system operations (creating files, updating directory structures), the file system itself can be left inconsistent, sometimes producing many corrupted files at once.
• Repeated crash damage to drives: physical drives subjected to repeated hard shutdowns from BSODs experience accelerated wear. Mechanical components have to repeatedly stop and restart suddenly.
• Cascade from underlying failure: if the BSOD was caused by a failing drive, continued operation while the user troubleshoots accelerates the failure. The drive may go from “occasional BSODs” to “completely unresponsive” over a period of hours or days.

The BSOD itself doesn’t damage the drive

This is the critical distinction users miss. The blue screen appearing is Windows protecting itself from a problem; the screen is not the problem. If the drive is healthy and the BSOD was caused by a software issue (driver bug, Windows update problem, application crash), data risk is minimal. If the drive itself is failing and producing the BSOD, the underlying failure is what threatens data, not the BSOD as such. The diagnostic question is which scenario applies, and the stop code is the primary clue.²

Every BSOD displays a stop code (sometimes called a bug check code) identifying the specific failure type. Several specific codes are direct signals that the storage drive is involved in the failure, while others point to issues that don’t typically affect data integrity.³

Stop code	What it means	Drive involvement
KERNEL_DATA_INPAGE_ERROR	Windows couldn’t read data from the page file or system data	Strong; bad sectors or read errors likely
INACCESSIBLE_BOOT_DEVICE	Windows can’t access the drive containing the OS	Very strong; drive controller or file system issue
UNMOUNTABLE_BOOT_VOLUME	Boot drive found but file system unreadable	Very strong; file system or media problem
CRITICAL_PROCESS_DIED	A core Windows process crashed	Possible; sometimes due to storage failure preventing process loading
FILE_SYSTEM_ERROR	Generic file system error	Strong; file system corruption
NTFS_FILE_SYSTEM	NTFS-specific file system error	Strong; NTFS structure damaged
FAT_FILE_SYSTEM	FAT-specific file system error	Strong; FAT structure damaged
MEMORY_MANAGEMENT	Memory subsystem error	Weak; usually RAM, sometimes related to virtual memory on disk
IRQL_NOT_LESS_OR_EQUAL	Driver or hardware error accessing memory	Weak; usually driver issue
PAGE_FAULT_IN_NONPAGED_AREA	Memory access error	Weak; usually RAM, sometimes paging file on disk

The “drive failure” cluster

The first six codes in the table above (KERNEL_DATA_INPAGE_ERROR, INACCESSIBLE_BOOT_DEVICE, UNMOUNTABLE_BOOT_VOLUME, CRITICAL_PROCESS_DIED, FILE_SYSTEM_ERROR, NTFS_FILE_SYSTEM) are the ones recovery engineers see most often associated with failing drives. If any of these appear, treat them as urgent warnings about drive health. The MDrepairs source notes that one BSOD with a drive-related stop code is worth investigating; most users wait until the tenth before acting, and by then the drive has been operating in a degraded state for weeks.

SSDs produce drive-related BSODs too

SSDs can fail more abruptly than hard drives, sometimes going from working perfectly to completely dead with little warning. The same stop codes appear: KERNEL_DATA_INPAGE_ERROR, UNMOUNTABLE_BOOT_VOLUME, CRITICAL_PROCESS_DIED. The underlying causes are different (worn NAND cells, controller failures, FTL corruption from firmware corruption) but the user-visible result is the same: Windows can’t read what it needs, so it crashes. SSD failures sometimes produce only one or two BSODs before the drive becomes entirely unresponsive, so SSD users have less warning time than HDD users.

Boot loops are when the system gets stuck in a cycle of attempting to boot, encountering a BSOD, restarting, and crashing again before completing the boot process. The system never reaches the desktop, and the user’s data is inaccessible through the normal interface.⁴

Why boot loops accelerate damage

Each cycle of the boot loop is a hard shutdown followed by a power-on. For mechanical drives, that means:

• The platters spin down rapidly under power loss rather than completing a controlled shutdown.
• The heads have to be parked (or land in the landing zone) without the normal procedural sequence.
• On startup, the heads have to be loaded back into operation under stress.
• If the original failure was bad sectors, each boot attempt drags the heads across the same problematic regions repeatedly.

For SSDs, repeated power cycles stress the drive’s controller and can corrupt firmware modules that were being updated when power was lost. A drive that was producing one BSOD per day under normal use can be entirely dead after several hours of boot looping.

Breaking out of the boot loop

The right move when a system enters a boot loop with drive-related stop codes is to power off completely and stop attempting to boot. Then:

• Disconnect power entirely (unplug the laptop, remove the battery if accessible, or disconnect the desktop’s power cable).
• Don’t keep retrying hoping the next attempt will succeed.
• Plan recovery from outside the failing system rather than from within it.

Distinct from a “non-bootable” drive without BSODs

A system that fails to boot without ever showing a BSOD is in a different category. The drive may still be readable; the boot record or system files may be the only damaged components. The standard recovery approach (drive removal or WinPE bootable media) usually works in those scenarios. BSOD-driven boot loops imply the drive is actively failing in ways the OS can detect, which means the recovery window is shorter and the urgency is higher.

Once the system can no longer boot reliably, the recovery approach moves outside the failing system. Two primary approaches exist, with different tradeoffs.⁵

Approach 1: Remove the drive, connect to a healthy computer

The fastest and most reliable approach when the user has access to a second computer. Steps:

1. Power off the failing computer completely.
2. Open the case (desktop) or access the drive bay (laptop).
3. Disconnect the drive from the SATA/NVMe interface and remove it.
4. Connect the drive to a healthy computer using a USB-to-SATA adapter, an NVMe USB enclosure, or as a secondary internal drive.
5. The healthy computer’s operating system should detect the drive and offer access to its files.
6. Copy critical data immediately to a separate destination drive.

This approach has key advantages: the failing system isn’t running, so no further BSODs occur during recovery; the working system isn’t dependent on the failing drive’s OS to boot; and standard file-copy tools work without specialized recovery software if the drive is readable.

Approach 2: Bootable WinPE recovery media

When physically removing the drive isn’t practical (warranty restrictions, soldered storage, sealed devices), bootable Windows Preinstallation Environment (WinPE) media provides an alternative:

1. On a healthy computer, create a bootable WinPE USB drive using tools like EaseUS Data Recovery Wizard’s “Recover from Crash Computer” feature, MiniTool’s bootable media builder, or Microsoft’s Windows Assessment and Deployment Kit (ADK).
2. Insert the USB into the failing computer and boot from it (typically by pressing F2, F12, or Esc during startup to reach boot menu).
3. The system boots into a minimal Windows environment running entirely from USB without touching the failing internal drive.
4. Use the included recovery software to scan the failing drive and copy files to an external destination.

This approach trades some complexity for accessibility. It works when drive removal isn’t an option but requires a working computer to create the bootable media in the first place.

When professional services are needed

Some BSOD-driven failures don’t yield to either DIY approach. Specifically:

• The drive isn’t detected on a different computer. If the failing drive doesn’t appear when connected to a healthy system, the issue is at the hardware or firmware level. See drive not recognized for the diagnostic ladder.
• The drive makes unusual sounds. Clicking, grinding, or scraping during boot attempts indicates physical damage. See click of death; stop attempting to boot and contact a professional service.
• SMART data shows critical errors. If the drive is detected on a healthy system but SMART attributes show rising bad-sector counts or imminent-failure flags, image the drive immediately and consider professional recovery for the most valuable data.
• The data value justifies the cost difference. Professional recovery is usually less risky than repeated DIY attempts on a failing drive.

Once data is safely backed up (or while it’s being backed up if the system is still partially functional), the underlying cause needs identification to prevent recurrence. The diagnostic approach varies by what’s accessible.⁶

If the system still boots intermittently

Capture diagnostic information during stable periods:

• Check Event Viewer: Windows logs the events leading up to each BSOD in System and Application logs. Check for repeated disk errors, driver crashes, or hardware events around the time of crashes.
• Run SMART tools: CrystalDiskInfo on Windows shows SMART attributes. Rising Reallocated Sectors, Pending Sectors, or Reported Uncorrectable Errors point to bad sectors developing.
• Check the page file location: KERNEL_DATA_INPAGE_ERROR specifically relates to page file reads. If the page file is on a failing drive, moving it to a different drive may stabilize the system temporarily.
• Note recent changes: recent driver updates, Windows updates, hardware additions, or applications installed before the crashes started point toward software causes.

If the system won’t boot at all

Diagnose from outside the system:

• Connect the drive to another computer and run SMART tools against it. If SMART shows critical errors, the drive is failing.
• Try a CHKDSK pass only if SMART data is clean and the file system needs repair. Don’t run CHKDSK on a drive showing critical SMART errors; it can make recovery harder.
• Use the Last Known Good Configuration option on older Windows versions if the issue started after a recent change.
• Try Windows Recovery Environment through Windows installation media; it offers options for system restore, startup repair, and command prompt access.

Distinguishing software from hardware causes

Indicator	Likely cause	Action
BSOD started after Windows update	Software (driver/update)	Roll back update or system restore
BSOD started after driver install	Software (driver)	Uninstall driver in safe mode
BSOD with drive-related stop code	Hardware (drive likely)	Back up data immediately
SMART data shows rising errors	Hardware (drive failing)	Replace drive after data backup
BSODs random with varying codes	Hardware (often RAM)	Run memtest86 to confirm
Drive makes unusual sounds	Hardware (physical drive failure)	Stop using; professional recovery
BSOD only in specific applications	Software (application or driver bug)	Update or replace application