Fix "A Disk Read Error Occurred" on Windows 10/11 (Exact Commands)

The fast answer. "A disk read error occurred. Press Ctrl+Alt+Del to restart" appears on a black screen before Windows loads — the firmware found the drive, but the early boot code couldn't read what it needed from it. Pressing Ctrl+Alt+Del just loops back to the same screen. Rule out hardware first: power off, reseat the drive's data and power cables, and confirm the drive still shows up in BIOS/UEFI setup. If the drive is detected, the most common software cause is a damaged file system or boot record. Boot a Windows 10/11 install USB, press Shift+F10 for a Command Prompt, and run:

chkdsk C: /f /r
bootrec /fixmbr
bootrec /fixboot
bootrec /rebuildbcd

Replace C: with the Windows drive letter shown in the recovery environment. Because a failing drive is a leading cause of this error, back up anything irreplaceable before running repairs.

Why this happens

Failing hard drive or SSD — bad sectors in the boot area, a failing controller, or a drive that responds intermittently. This is one of the most common causes, especially on older spinning drives, and often comes with clicking or grinding noises.
Loose or damaged cables — a SATA data or power cable that has worked loose or failed makes reads fail before Windows ever starts.
Corrupted MBR, boot sector, or partition table — disk write errors, power loss during a write, or malware damaged the boot records the firmware hands control to.
File-system corruption — an unclean shutdown left the boot volume in a state the loader can't read.
Bad clone or partition change — the error commonly appears right after cloning to a new drive or resizing partitions, when boot records don't match the new layout.
Faulty or reseated RAM — a well-documented oddball cause; failed memory can corrupt reads during early boot.
Wrong boot device or BIOS settings — the firmware is trying to boot from a disk that has no bootable system on it, often after a CMOS reset.

Step-by-step fix

Step 1 — Reseat cables and confirm the drive is detected

Power off and unplug the machine. Unplug any USB drives and external disks. On a desktop, reseat both ends of the SATA data cable and the power connector on the boot drive — swap in a known-good SATA cable if you have one. Then enter BIOS/UEFI setup (usually Del or F2 at power-on) and confirm the drive appears by name. If the drive is missing from the BIOS, or appears only sometimes, this is a hardware problem — no command fixes a drive the firmware can't see. Stop and prioritize data recovery.

Step 2 — Check the boot order

While in BIOS/UEFI setup, check the boot order. After a CMOS reset or battery failure, the firmware can try to boot a data disk or an empty drive first. Put the Windows drive at the top of the list, save, and reboot.

Step 3 — Test the RAM

Failed or badly seated memory is a documented cause of this exact error. Power off, remove the RAM modules, and reseat them firmly. If you have two or more sticks, try booting with one at a time in the first slot. If the error follows one particular stick, replace it. Once Windows boots again, run Windows Memory Diagnostic (mdsched) for a fuller test.

Step 4 — Run chkdsk from a Windows install USB

This error appears before Windows starts loading, so forcing three failed boots usually can't trigger Automatic Repair — use a Windows 10/11 install USB instead. Boot it, and at the first setup screen press Shift+F10 to open Command Prompt (or choose Repair your computer → Troubleshoot → Command Prompt). Confirm the Windows drive letter first — it may not be C: inside the recovery environment — then run:

chkdsk C: /f /r

/f fixes file-system errors and /r locates bad sectors and recovers readable data. On a large or damaged drive this can take hours — let it finish. A long list of bad sectors means the drive is failing: back up and replace it rather than reusing it.

Step 5 — Repair the MBR and boot records

From the same Command Prompt, run these in order:

bootrec /fixmbr
bootrec /fixboot
bootrec /rebuildbcd

When bootrec /rebuildbcd finds a Windows installation, type Y to add it. Reboot and test.

UEFI/GPT note: on modern UEFI systems bootrec /fixboot often returns "Access is denied" — that's normal there and not the real fix. Rebuild the EFI boot files with bcdboot instead: assign a letter to the EFI System Partition with diskpart, then run bcdboot C:\Windows /s S: /f UEFI. The full walkthrough is on the 0xc000000e page.

Step 6 — Reset BIOS to defaults, then consider an update

If the error persists, load the BIOS/UEFI defaults ("Load Setup Defaults" or similar) from firmware setup, save, and reboot — this clears a misconfigured drive-mode or boot setting you may not spot by eye. If defaults don't help, a BIOS update from the motherboard or PC maker's site has resolved this error on some systems, but only attempt it on stable power and by following the vendor's procedure exactly — a failed flash is worse than the error you started with.

Step 7 — If it started after a clone, redo the clone

If the error appeared the first time you booted a freshly cloned drive, the clone itself is the likely problem: boot records or partition layout didn't survive the copy. Boot the install USB and run the Step 5 repairs against the new drive. If it still fails, redo the clone with the source drive connected and healthy, or do a clean Windows install on the new drive and copy your files over afterward.

Data safety: Treat this error as a possible drive failure until proven otherwise. If the files on this machine matter and you don't have a backup, image the drive to a known-good external disk before you run repairs — chkdsk /r stresses weak sectors, and a drive that's already dying can degrade further or fail completely under that load. A clicking or grinding drive should go straight to data recovery, not repair. Never run repairs on the only copy of irreplaceable data.

GRAM tip: Once the machine boots again, plug in GRAM — a free portable repair toolkit that runs from a USB inside Windows — to read the drive's SMART health (critical here: reallocated and pending sectors tell you whether the drive that threw this error is dying), check the event logs, and let its AI agent confirm whether the real cause was the drive, a cable, RAM, or damaged boot records so the error doesn't come back.

Download GRAM free See AI pricing →

FAQ

Can I fix this without a second working computer?

Often, yes. Cable reseating, BIOS boot-order checks, and RAM swaps need no media at all. But note one difference from blue-screen errors: this message appears before Windows starts loading, so forcing three failed boots usually can't trigger Automatic Repair. For the chkdsk and bootrec repairs you'll need a Windows 10/11 install USB — any existing one works, though creating a fresh one does require a working PC. (GRAM runs inside Windows, not from boot media — its role comes once the machine boots again: verifying disk health and confirming the root cause.)

Does "A disk read error occurred" mean my hard drive is dead?

Not necessarily, but take the possibility seriously — a failing drive is one of the most common causes. Loose cables, corrupted boot records, a bad clone, and even faulty RAM produce the identical message. Work through the cable, BIOS, and RAM checks first; if chkdsk /r reports many bad sectors, or the drive clicks or disappears from the BIOS intermittently, the drive is failing — back up immediately and replace it.

Will chkdsk or bootrec erase my files?

No. bootrec rewrites only boot records, not your data, and chkdsk /f /r repairs the file system — it can move damaged file fragments to FOUND.000 folders but does not wipe the drive. The real risk is running chkdsk /r on a drive that's already failing, because the sector scan stresses weak areas. If the data is irreplaceable, image the drive first.

Why did this appear right after cloning to a new SSD?

Cloning is a classic trigger. If the boot records or partition table didn't transfer cleanly, the new drive's early boot code can't read what it expects and throws this error. Run the bootrec sequence (or bcdboot on UEFI systems) against the new drive from an install USB. If that doesn't fix it, redo the clone — and if the source drive shows read errors during cloning, that's the failing part, not the new SSD.