Loading
Main Pages
Home About Contact
Diagnostic Guides
HDD Health & Diagnostics RAM Testing Methods CPU Temperature Monitoring POST Error Codes Explained Reading SMART Data
Legal
Privacy Policy

Reading SMART Data: A Complete Guide to Drive Health Attributes

SMART data is the most direct window into a drive's internal health. Understanding what each attribute means — and which ones actually predict failure — turns a wall of numbers into actionable information.

Plextor M8Pe NVMe SSD 256GB

How SMART Works

SMART (Self-Monitoring, Analysis and Reporting Technology) is a monitoring system built into virtually all modern storage devices — HDDs, SATA SSDs, and NVMe drives. The drive's firmware continuously tracks dozens of internal metrics and stores them in a dedicated area of the drive's memory.

These metrics are accessible to the operating system and diagnostic software. The drive also maintains threshold values for each attribute. When a raw value crosses its threshold, the drive sets a health flag that tools like CrystalDiskInfo display as a Caution or Bad status.

Important Context

SMART thresholds are set conservatively by manufacturers. A drive can show a Caution status and continue working reliably for months. Conversely, some drives fail without any SMART warning. SMART is a useful indicator, not a guarantee.

How to Read SMART Data

Download and install CrystalDiskInfo. Open it and select your drive from the tabs at the top. The attribute table shows:

  • ID — Hexadecimal attribute identifier
  • Current — Normalized value (higher is generally better, scale varies)
  • Worst — Lowest Current value ever recorded
  • Threshold — Minimum acceptable Current value set by the manufacturer
  • Raw Value — The actual measured data in hexadecimal

For most attributes, the Raw Value is what you actually care about. The Current/Worst/Threshold system is normalized and can be confusing — focus on the raw numbers for the critical attributes below.

Critical HDD Attributes

These are the attributes with the strongest correlation to imminent drive failure, based on large-scale studies including Backblaze's drive reliability research:

ID 05 — Reallocated Sectors Count

The number of sectors that have been moved to a reserved spare area because they became unreadable or unreliable. Each reallocation means the drive found a physical problem and worked around it. Any non-zero raw value is a warning. A rising value over time is a strong predictor of failure.

ID C5 — Current Pending Sector Count

Sectors that are currently unstable and waiting to be reallocated or verified. These sectors cannot be read reliably right now. A non-zero value means the drive is actively struggling. If the drive successfully reads these sectors on a subsequent attempt, the count may decrease.

ID C6 — Uncorrectable Sector Count

Sectors that could not be read or corrected by the drive's error correction system. Even a value of 1 is serious. This attribute directly indicates data loss or corruption.

ID 01 — Raw Read Error Rate

The rate at which errors occur when reading data from the disk surface. Seagate drives often show large raw values here by design — this is a known quirk and does not indicate a problem on its own. For other manufacturers, a high and rising value is a concern.

ID 0A — Spin Retry Count

How many times the drive had to retry spinning up to operating speed. Any non-zero value suggests mechanical issues with the spindle motor or bearings.

ID BC — Command Timeout

Operations that timed out while waiting for the drive to respond. Elevated values indicate the drive is struggling to complete basic operations in time.

Important SSD Attributes

SSDs use different attributes because they have no moving parts. The relevant metrics reflect NAND flash wear rather than mechanical degradation.

ID E8 / E9 — Available Reserved Space / Media Wearout Indicator

Indicates how much of the drive's reserved spare area remains. As NAND cells wear out, the drive uses this reserve to maintain performance and reliability. When it approaches zero, the drive is near the end of its rated write endurance.

ID F1 — Total LBAs Written

The total amount of data written to the drive since manufacture. Compare this to the drive's rated TBW (Terabytes Written) specification to gauge remaining lifespan.

ID 05 — Reallocated Sectors (SSDs)

SSDs also track bad blocks. Unlike HDDs, a small number of reallocated blocks is expected over time. What matters is whether the count is stable or rising.

Inside view of hard disk drive platters
HDD platters store data magnetically. SMART tracks the health of these surfaces through attributes like Reallocated Sectors. Source: Wikimedia Commons

NVMe SMART Data

NVMe drives report SMART data through a different interface than SATA drives, but CrystalDiskInfo reads both. Key NVMe attributes include:

  • Percentage Used — Estimated percentage of rated write endurance consumed (0–100%)
  • Available Spare — Percentage of spare capacity remaining
  • Data Units Written — Total data written in 512KB units
  • Media and Data Integrity Errors — Errors detected by the NVMe controller. Should be zero.

Monitoring Over Time

A single SMART reading is less informative than tracking values over time. CrystalDiskInfo can be configured to run at startup and log attribute values. What you are looking for is change — specifically, whether critical attributes like Reallocated Sectors are stable or increasing.

A drive with 10 reallocated sectors that has been stable for two years is less concerning than a drive that went from 0 to 5 reallocated sectors in the past month.

Research Reference

The most comprehensive public data on SMART attribute predictive value comes from Backblaze's analysis of over 100,000 drives. Their research confirms that attributes 05, C5, C6, and BC are the strongest failure predictors.