Why an SSD Can Fail at 100% Health

An SSD can report 100% health and still fail. In many utilities, that percentage estimates how much of the drive's rated flash endurance remains; it is not a complete diagnosis of the controller, firmware, circuitry or data integrity. Treat it as one clue, not a guarantee.

If a drive reports a critical warning, becomes read-only, disappears intermittently or begins logging media errors, back up important files immediately. Do not wait for the percentage to fall to an arbitrary number.

What the health percentage actually measures

NAND flash cells tolerate a limited number of program-and-erase cycles. SSD makers specify endurance in figures such as terabytes written (TBW), and an NVMe drive can expose a SMART value called Percentage Used. A value of 0 means the drive has used essentially none of its rated endurance; it rises as estimated wear accumulates. Some programs invert that number and display the result as “health remaining.”

This estimate does not cover every failure mode. A drive with little flash wear can still develop a controller fault, firmware problem, damaged connector, power-related failure or uncorrectable data error. Conversely, reaching the rated endurance does not mean the SSD will stop at that exact moment—it means the manufacturer-defined endurance has been consumed and replacement planning is prudent.

SSD utility showing drive health and SMART information

SSD warning signs that matter

Signal	What it can mean	What to do
NVMe Critical Warning	Low spare capacity, degraded reliability, excessive temperature, read-only mode or another controller-reported problem	Back up at once; check the manufacturer's diagnostic and plan replacement
Available Spare below threshold	The reserve flash used to replace unusable blocks is running low	Treat it as a serious warning, not normal wear
Media and Data Integrity Errors	The controller could not recover some data	Watch for any increase and replace the drive if errors are present
Uncorrectable errors or file corruption	Reads or writes are no longer reliably completing	Stop nonessential writes and secure a backup
Drive repeatedly disappears or turns read-only	Possible controller, connection, power or media failure	Check the connection once, then prioritize data recovery and replacement
High temperature	Poor cooling can cause throttling and may reduce reliability	Improve airflow and verify the temperature under load

Counts such as Unsafe Shutdowns and Error Information Log Entries need context. An unsafe-shutdown count can rise after a power loss or forced restart and does not by itself prove the SSD is failing. A cumulative error-log count is most useful when paired with the error type, operating symptoms and whether the value continues to increase.

How to check an SSD in Windows

Protect the data first. If Windows warns that storage reliability is degraded, or files are already unreadable, copy irreplaceable data before running lengthy tests. A second copy that can be restored is more valuable than another health scan; see the TipsMake guide to creating a resilient backup.
Check Windows storage health. In Windows 11, open Settings > System > Storage > Advanced storage settings > Disks & volumes, select the drive and open its properties. Windows can display estimated life, available spare and temperature for supported NVMe drives. No warning here does not prove every SATA or USB drive is healthy.
Read the complete SMART/NVMe data. CrystalDiskInfo is a convenient option. Select the correct physical drive, note its model and firmware, then inspect Critical Warning, Available Spare, Percentage Used, Media and Data Integrity Errors and temperature rather than relying only on the colored status badge.
Compare with the maker's tool. Samsung, Crucial, Western Digital and other vendors provide diagnostics or firmware utilities for supported models. Attribute names and raw-value formats differ, especially on SATA SSDs, so the manufacturer's interpretation may be clearer.
Check other likely causes. A nearly full SSD can slow down without being defective. Faulty cables, unstable power, overheating, memory errors and corrupted system files can also resemble drive failure. TipsMake's guide to checking storage pressure helps separate capacity problems from hardware warnings.

Detailed SMART and NVMe values used to assess an SSD

When should you replace the SSD?

There is no universal “replace at 70% health” rule. Replace or retire a drive promptly when it reports a critical warning, low spare capacity, media errors, read-only mode or repeated unexplained disconnects. Increasing error counts and recurring file corruption are also stronger reasons to act than a wear percentage alone.

If the only issue is a steadily rising Percentage Used value and the drive has no errors, use that trend to budget and schedule a replacement. For important systems, replace before reliability becomes uncertain and test the restore process rather than assuming a backup works.

SMART monitoring is useful, but it cannot predict every sudden electronics or controller failure. The dependable safeguard is a current backup stored separately from the SSD.