Programs take forever to open, the PC freezes for no clear reason, files go missing, or Windows itself pops up a disk warning. Hard drives and SSDs rarely die silently — they keep score of their own decay in SMART data. The trick is reading that score in time, and a WinPE toolkit with CrystalDiskInfo reads it even when Windows barely runs.
🎥 The toolkit USB carrying CrystalDiskInfo and the other diagnostics is built step by step in the video
Treat these as the classic early warnings of a dying drive: long freezes while opening files, clicking or grinding sounds from an HDD, files and folders that vanish or refuse to open, painfully slow copies, and Windows repair loops. None of them proves a bad disk on its own — but together they demand a health check before anything else, because every other repair (reinstalling Windows, running heavy scans) is wasted work, and extra stress, on a disk that is physically leaving. The check itself is free, read-only and finished in minutes.
WINDOWS DIALOG — THE WARNING WINDOWS ITSELF SHOWS
Microsoft Windows✕
⚠ Windows detected a hard disk problem
Back up your files immediately to prevent information loss, and then contact the computer manufacturer to determine if the disk needs to be repaired or replaced.
Start the backup processAsk me again later
If you have seen this dialog even once, believe it. It appears only after SMART has already reported serious trouble.
📷 What Windows' own disk warning looks like — this dialog is never a false alarm to be dismissed
1
Boot WinPE and Open CrystalDiskInfo
Boot your Ventoy toolkit USB and choose Sergei Strelec WinPE or MediCat — both carry CrystalDiskInfo in their diagnostics sections (it also runs inside normal Windows whenever Windows still starts)
Open CrystalDiskInfo and wait a few seconds while it reads every attached drive
If the PC has several drives, use the disk tabs/menu at the top to check each one — the slow symptoms only need one bad disk among them
2
Read the Verdict: Health Status and the Three Key Numbers ⭐
Start with the big verdict on the left: Good, Caution or Bad. On an SSD it is shown as a percentage — that figure is the drive's remaining wear life
Then read the Raw Values column for three attributes: 05 Reallocated Sectors Count, C5 Current Pending Sector Count and C6 Uncorrectable Sector Count. On a healthy drive all three raw values are 0
Any non-zero, rising value means the disk surface is already damaged and the drive is quietly working around dead areas
Glance at Temperature and Power On Hours for context — a very hot or extremely old drive deserves extra suspicion
CRYSTALDISKINFO STYLE REPORT — A DRIVE IN TROUBLE
Health Status⚠ Caution
05 — Reallocated Sectors Count (raw)200
C5 — Current Pending Sector Count (raw)48
C6 — Uncorrectable Sector Count (raw)16
Temperature43 °C
Power On Hours38,412 hours
📷 What to look at — the status word, plus raw values that should be zero on a healthy drive
CrystalDiskInfo and its companion diagnostics ride on the same toolkit USB built on screen in the video above.
3
Cross-Check With Hard Disk Sentinel
Both toolkits also include Hard Disk Sentinel, which reads the same SMART data and phrases it as a plain-language health percentage and verdict. Two tools agreeing removes doubt. One discipline rule: on a drive already suspected of dying, do not start a full surface scan before backing up — a surface test pounds every sector for hours, which is exactly the stress a failing drive may not survive.
4
Act on the Result — In This Order
Caution or Bad: stop troubleshooting Windows. Rescue your files now — copy them from WinPE to an external drive (see this site's data rescue guide) or clone/image the whole disk to a new one with the toolkit's imaging tools
Replace the drive. A disk with growing reallocated sectors does not heal, and "it still works today" is not a plan
Good health, symptoms persist: believe the result and look elsewhere — test the RAM with MemTest86+, and check Windows itself. Not every slow PC has a dying disk
Whatever the result, note the numbers. Checking again in a month and seeing the counts rise is the clearest possible evidence of decline
🛟 SMART is a warning system, not a guarantee: a drive showing "Good" can still fail suddenly from an electronic fault — SMART mainly tracks gradual wear and surface decay. The health check tells you when to run; the only real protection, at any health score, is a backup that already exists.
🎥 Build the Diagnostic Toolkit in Video
Ventoy + Strelec + MediCat on one USB, diagnostics included — the complete build is shown step by step.
No. For an SSD that number is remaining wear life, and 90% simply means about a tenth of its rated write endurance is used — normal for a drive in service. What deserves action is a low percentage, a fast fall between checks, or a Caution/Bad status alongside it.
CrystalDiskInfo does not show my drive when it is connected through a USB adapter. Why?
Many USB-to-SATA adapters and cheap enclosures do not pass SMART data through, so health tools see a disk with no report. Connect the drive directly — SATA cable or the NVMe slot inside the PC — to read its health properly.
The status says Caution, but the PC works fine. Should I really replace the drive?
Yes — on your schedule rather than the drive's. Caution means physical damage has already begun (typically reallocated sectors). Such drives sometimes run for months and sometimes die this week; the status cannot tell you which. Back up immediately and replace the drive while copying is still easy.
This guide is based on Technical Anshu's "One USB = All Tools" video • Stuck on a step? Ask in the video comments — you'll get help there 🙏