Triage a Windows no-boot case when no cause or installation is yet detected

This is a routing procedure for a no-boot case that does not yet contain enough evidence for a targeted repair. It also covers Total identified Windows installations: 0 from bootrec /scanos or /rebuildbcd. That result does not prove Windows was erased and does not automatically mean BCD is corrupt. The Windows volume might be locked, missing from the storage path, unlettered, damaged, an unsupported file system, or simply not recognized by Bootrec’s scan.

Searchable Windows error codes and exact messages

Use this when the available report says only “won’t boot,” “repair loop,” “black screen,” or “Windows installations: 0,” and no specific mechanism is yet demonstrated.

Once a named code, file, driver, partition defect, hardware failure, or boot stage is confirmed, leave this router and use the corresponding targeted fix-it. Do not continue broad triage after the cause is established.

The exact build, original trigger, and final cause may remain unknown during the first pass. The exact visible symptom, boot stage, physical-disk visibility, offline Windows volume, and storage health may not remain unknown before any write.

• Current Windows installation/recovery media
• Phone/camera or written transcript for exact messages
• Administrative Command Prompt
• Backup destination when logs or readable files must be preserved
• No destructive disk operation

Disconnect unrelated USB storage and memory cards. Enter Repair your computer > Troubleshoot > Advanced options > Command Prompt. Do not run a memorized sequence of bootrec, chkdsk /f /r, registry-copy, or partition commands.

1. Classify the last visible stage:

   • before Windows Boot Manager: no device, firmware shell, PXE, blinking cursor;
   • Windows Recovery: code, File:, and Info:;
   • Windows logo/spinner;
   • blue-screen stop code;
   • sign-in screen;
   • post-login black screen.

2. Record every connected disk from firmware, then compare in WinRE:

   diskpart
   list disk
   list volume
   exit
   

3. If the system disk is absent in both firmware and WinRE, stop Windows repair and diagnose hardware/power/controller enumeration.

4. If firmware sees it but WinRE does not, investigate the storage driver/controller path before BCD.

5. If volumes are visible, examine candidate volumes without assigning familiar letters by guess. Select by size/file system, assign an unused temporary letter, and inspect:

   diskpart
   select volume <candidate-number>
   detail volume
   assign letter=W
   exit
   

   dir W:\Windows\System32\Config\SYSTEM
   dir W:\Windows\System32\winload.efi
   dir W:\Windows\System32\winload.exe
   

6. A candidate is a plausible Windows installation when the directory tree and core files agree. bootrec /scanos returning zero does not override direct evidence.

7. Run read-only checks first:

   chkdsk W:
   dism /Image:W:\ /Get-CurrentEdition
   

8. Copy available recovery evidence before changes:

   type W:\Windows\System32\Logfiles\Srt\SrtTrail.txt
   dir W:\Windows\Logs\CBS
   dir W:\$WINDOWS.~BT\Sources\Panther
   

9. If BCD is readable, export it before any modification. For the live store used by the booted environment:

   bcdedit /enum all /v
   bcdedit /export W:\BCD-backup-<date>.bcd
   

   If an offline store is being inspected, use /store <full-path> and copy that exact file separately.

1. Route by the first decisive evidence:
   • missing disk in firmware: hardware/storage-enumeration fix;
   • disk appears only after driver load: FIX-051;
   • 0x7B after controller change: controller-mode fix;
   • missing/corrupt BCD 0xc000000f: FIX-004;
   • required device 0xc0000225: FIX-011;
   • invalid OS information 0xc0000098: FIX-012;
   • named winload.efi: FIX-013;
   • SYSTEM hive 0xc000014c: FIX-017;
   • SrtTrail.txt without isolated cause: FIX-026;
   • disk/file-system errors: FIX-042;
   • no bootable device with intact Windows: FIX-005.
2. If Windows is directly verified but Bootrec still reports zero, do not force Bootrec to “find” it. Inspect the actual system partition and use the targeted UEFI/GPT or BIOS/MBR boot-file procedure only after partition style and system partition are proven.
3. If no candidate contains a coherent Windows directory, search all readable NTFS volumes and check whether the disk/partition map changed. Preserve data before attempting partition recovery.
4. If the volume is RAW, unreadable, or produces I/O errors, stop boot repair and image/recover data first.
5. Perform only the selected targeted repair, then test before adding another change.

The intended result is a specific diagnosis, not necessarily an immediate boot. A good first pass narrows the case to one boot layer and preserves evidence. If no cause emerges, retain it as unclassified rather than claiming BCD repair succeeded.

Do not run diskpart clean, initialize a disk, format an “unknown” partition, copy RegBack blindly, mark partitions Active on GPT, or execute every bootrec command because one scan returned zero. Do not reinstall until readable user data and the existing layout are preserved.

Read-only triage requires no rollback. Remove temporary letters if desired:

diskpart
select volume <volume-number>
remove letter=W
exit

Rollback any later targeted repair using that article’s instructions and the saved BCD/image.

Verify that the physical disk map, partition map, Windows volume, exact screen, and next repair route are documented. If a repair was applied, verify repeated boots and recheck the evidence that justified it.

Confidence is about 80% in the routing decision when two independent observations agree—for example, firmware/WinRE visibility plus driver response, or an exact named file plus offline inspection. Stop for ambiguous disks, RAW volumes, I/O errors, missing encryption credentials, or conflicting evidence.