To check a hard drive for bad sectors on Windows, you have two options: read the drive's SMART data, which shows how many bad sectors the drive itself has already found, and run a surface scan, which reads every block on the disk to find new ones. Windows can do a surface scan with chkdsk /r, but it also writes repairs to the file system as it goes. A read-only surface scan is the safer first step, especially on a drive that may be failing, because it finds problems without changing anything.
This guide explains what bad sectors are, how to check for them with built-in Windows tools, how to read the results, and what to do next.
What bad sectors are
A hard drive stores data in small blocks called sectors, usually 4,096 bytes each on modern drives. A bad sector is one the drive can no longer read or write reliably. There are two kinds:
- Physical (hard) bad sectors are damaged areas of the disk surface, caused by wear, a manufacturing flaw, or the read head touching the platter. They cannot be repaired, only avoided.
- Logical (soft) bad sectors contain data that does not match its error-correction code, often after a power cut during a write. The surface is fine, and the sector becomes usable again once it is rewritten.
Every drive ships with spare sectors. When it finds a sector that is truly bad, it quietly swaps in a spare and records the event in its SMART log as a reallocated sector. You only start to see errors in Windows when the drive cannot read a sector at all, or when it runs short of spares.
Step 1: Check SMART data first
Before scanning, look at what the drive already knows about itself. The three SMART values that track bad sectors are:
| SMART attribute | ID | What it counts |
|---|---|---|
| Reallocated Sectors Count | 5 | Bad sectors already replaced with spares |
| Current Pending Sector Count | 197 | Unreadable sectors waiting to be dealt with |
| Offline Uncorrectable | 198 | Sectors that failed the drive's own offline test |
If any of these are above zero, back up before you run a long scan. A full surface scan reads the entire disk, and a drive that is already struggling may get worse under that workload. For what the other values mean, see SMART attributes explained.
Step 2: Scan with chkdsk
Windows has a built-in bad sector scan in the chkdsk command. Use the /r switch, which finds bad sectors, recovers readable data from them and marks them so the file system will not use them again. It includes everything /f does, so it also fixes file system errors.
- Right-click Start and choose Terminal (Admin) or Command Prompt (Admin).
- Type
chkdsk D: /r, replacing D: with the drive you want to check, and press Enter. - If Windows says the drive is in use, type Y to dismount it, or to schedule the check for the next restart if it is your C: drive.
- Wait. On a large hard drive this can take several hours. Do not turn off the PC while it runs.
When it finishes, look for a line such as "0 KB in bad sectors". Any value above zero means chkdsk found and marked bad areas. Our guide on how to run chkdsk in Windows 11 covers the other switches and how to find the results after a restart scan.
The Error checking button in a drive's Properties → Tools tab checks the file system only. It does not read every sector, so it will not find bad sectors in unused space.
Why a read-only scan is safer
chkdsk /r is a repair tool. When it finds damage, it changes the file system: it may move data, truncate files it cannot read, or put orphaned fragments in a folder named found.000. On a healthy drive with a few soft errors, that is what you want. On a drive that is failing, you want your files copied off first, and a long read-write repair is the wrong order of operations.
A read-only surface scan reads every block and reports which are bad or slow, without writing anything. You get the full picture, then decide what to do. Disk Detector runs this kind of scan with a live map of the drive as it goes, so you can see exactly where bad and slow areas are. Because it is read-only, nothing on the drive is changed.
How long a scan takes
A full surface scan reads the whole disk, so its speed is roughly the drive's sequential read speed. A typical hard drive reads at 120 to 250 MB/s, which works out at about 1.5 to 3 hours per terabyte. Scans slow down sharply when they hit damaged areas, because the drive retries each weak sector several times. SSDs scan far faster.
Reading the results
| Result | What it means | What to do |
|---|---|---|
| No bad or slow blocks, SMART counts at zero | The surface is healthy | Nothing; check again in a few months |
| A few slow blocks, no bad ones | Weak areas that still read | Back up and recheck in a month |
| A handful of bad blocks, stable SMART count | Some damage that is not spreading | Back up and monitor closely |
| Bad blocks clustered or growing between scans | The surface is deteriorating | Copy your data off and replace the drive |
The trend matters more than the number. Run a second scan or check SMART again a week later. If the count has grown, the drive is on its way out.
Checking an SSD for bad blocks
SSDs manage bad blocks internally and hide them from Windows, so a surface scan is less informative than on a hard drive. For an SSD, the more useful checks are the SMART values for media errors, available spare and percentage of life used. A read scan can still find blocks the SSD can no longer read, which is a serious sign. Disk Detector reads both kinds of drive and explains SSD wear values alongside hard drive sector counts.
What to do if you find bad sectors
- Back up first. Copy irreplaceable files before anything else. Our guide on how to back up a failing hard drive shows how to do it without stressing the drive.
- Run chkdsk /r once your files are safe, so the file system stops using the bad areas.
- Recheck in a week. Compare SMART counts and scan results.
- Replace the drive if the counts are rising. Bad sectors on a hard drive are not repaired, only worked around, and a growing number means more are coming.
Frequently asked questions
Is it safe to scan a drive for bad sectors?
A read-only surface scan is safe because nothing is written. chkdsk /r also changes the file system when it finds problems, so back up important files first.
How many bad sectors are too many?
Any bad sectors on a hard drive are a warning, but a number that keeps growing between scans is the real red flag. Replace a drive whose count is rising.
How long does a bad sector scan take?
A full read of a hard drive runs at roughly its sequential speed, so a 2 TB drive often takes 3 to 5 hours, and longer if it hits many slow areas.
