CrystalDiskInfo is the standard free tool for checking SSD health because it reads the SMART (Self-Monitoring, Analysis, and Reporting Technology) data every modern drive already tracks internally and presents it as a simple color-coded status: blue for good, yellow for caution, red for a drive that needs immediate attention. Checking a handful of key attributes, rather than trying to interpret every row, is enough to catch a genuinely failing drive well before it causes data loss.
Top 3 picks at a glance
Quick answer
Install CrystalDiskInfo, open it, and look at the overall health status color for each drive listed. Blue means healthy, no action needed. Yellow means back up your data now and watch the drive closely. Red means stop trusting that drive with anything important and plan a replacement. For deeper detail, the Reallocated Sectors Count and Percentage Used attributes are the two most worth understanding beyond the color status.
As an Amazon Associate we earn from qualifying purchases at no extra cost to you. Product prices and availability are accurate as of the date shown and are subject to change.
| Status color | What it means | Action |
|---|---|---|
| Blue (Good) | SMART attributes within normal range | None needed |
| Yellow (Caution) | One or more attributes near threshold | Back up data, monitor closely |
| Red (Bad) | Attribute has crossed failure threshold | Replace drive, stop relying on it |
| Percentage Used near 100% | NVMe endurance nearly consumed | Plan replacement before failure |
Why SSD health checking matters more than people expect
Solid state drives fail differently from mechanical hard drives. A failing hard drive often gives audible warning, clicking or grinding sounds, before it dies completely. An SSD gives no physical warning at all; it can go from working normally to completely unreadable within a single power cycle once its controller or NAND reaches a critical failure point. SMART monitoring is effectively the only early warning system available for these drives.
Unlike CPU or GPU temperature, which recovers the moment load decreases, SSD wear is cumulative and permanent. Every NAND flash cell has a finite number of program-erase cycles before it can no longer reliably hold data, and while modern drives are rated for enough total writes that normal gaming and everyday use rarely approaches the limit, heavy workloads like video editing scratch disks or constant large file transfers can wear a drive down meaningfully faster than typical use.
The stakes of an SSD failure are also different from most other component failures. A failed GPU or PSU is an inconvenience and a cost; a failed boot or storage drive can mean losing save files, screenshots, projects, or anything else that was not backed up elsewhere. Checking SSD health periodically, alongside the temperature checks covered in our temperature monitoring guide, is one of the cheapest forms of insurance available for a gaming PC.
Warranty claims also benefit from having SMART data on hand. Most SSD manufacturers accept a CrystalDiskInfo screenshot or exported report as supporting evidence for an RMA claim showing a drive has genuinely failed rather than simply stopped being recognized due to a cable or connection issue.
Installing and reading CrystalDiskInfo
CrystalDiskInfo is available from its official GitHub releases page or through the Microsoft Store; either the “standard” or “shizuku edition” (a themed variant with the same underlying data) works fine, the difference is purely cosmetic. On first launch, it automatically detects every drive connected to the system, including SATA SSDs, NVMe drives, and traditional hard drives, listing each with its own status panel.
The main window shows the drive model, firmware version, total capacity, interface type, and a large health status indicator at the top. Below that sits a scrollable table of individual SMART attributes, each with a raw value, a normalized value (typically on a 1-100 or similar scale depending on manufacturer), and a worst-recorded value.
Most users do not need to understand every attribute in that table. The handful worth knowing are covered in the following sections; the rest exist mainly for advanced diagnostics or manufacturer-specific troubleshooting, and their exact meaning varies enough between drive brands that broad generalizations about them are not reliable.
The software also supports a background monitoring mode that can alert you if a drive’s status changes, and can be set to launch at Windows startup with a minimized icon in the system tray, which is a reasonable way to get passive protection without needing to remember to open it manually.
Key SMART attributes explained
Reallocated Sectors Count tracks how many storage sectors have been marked as unusable and swapped for spare sectors reserved by the drive for exactly this purpose. On a healthy drive this value should be zero or stay flat over time; SATA drives with mechanical origins in their SMART standard show this attribute more prominently, but SSDs report an equivalent NAND-block-level metric under similar naming.
Percentage Used, found on most NVMe drives, is a direct estimate of how much of the drive’s total rated write endurance has been consumed, expressed as a simple 0-100% figure. This is arguably the single most useful attribute for gauging remaining SSD lifespan, since it is derived directly from the manufacturer’s own endurance rating for that specific model rather than a generic estimate.
Power-On Hours and Power Cycle Count give context rather than a direct health signal; a drive with very high power-on hours but a clean health status is simply an old, well-behaved drive, not necessarily one at risk. These numbers are more useful for confirming a used or refurbished drive’s history than for judging a drive you have owned since new.
Uncorrectable Error Count and Pending Sector Count are two of the more serious attributes to watch; any nonzero and climbing value here suggests the drive’s controller is encountering data it cannot reliably read or write, which is a stronger signal of imminent failure than most other attributes on the list.
NVMe versus SATA SSD health reporting differences
NVMe drives generally expose more detailed and more standardized SMART data than SATA SSDs because the NVMe specification defines a dedicated health log page that all compliant drives must support, while SATA SSDs inherited an older SMART standard originally designed for spinning hard drives and had to adapt attributes to fit. This is why Percentage Used shows up cleanly on most NVMe drives but is less consistently present on SATA SSDs.
NVMe drives also report a composite health status through their controller more directly, which is part of why CrystalDiskInfo’s overall color status tends to be more immediately meaningful on NVMe drives than on some older or budget SATA SSDs, where individual manufacturers have historically implemented SMART attributes with varying levels of completeness.
Our comparison of NVMe versus SATA SSDs for gaming covers the performance differences in more depth, but from a health-monitoring standpoint alone, NVMe’s more standardized reporting is a modest additional point in its favor for anyone who wants clearer long-term visibility into drive condition.
Regardless of interface, both types benefit from the same basic monitoring habit: check periodically, note the baseline when the drive is new, and watch for attributes that move meaningfully away from that baseline rather than fixating on a single snapshot in isolation.
Checking health on a used or second-hand drive
Buying a used SSD, whether individually or as part of a used prebuilt system, makes a SMART check one of the first things worth doing before trusting it with anything important, since power-on hours and total bytes written give a much clearer picture of a used drive’s actual history than its outward condition or the seller’s description alone. A drive advertised as “lightly used” with genuinely low power-on hours and a healthy Percentage Used figure is a meaningfully safer purchase than one where those numbers can’t be confirmed.
Reallocated sector count and any nonzero uncorrectable error count are particularly worth scrutinizing on a used drive, since these attributes reflect accumulated wear and potential defects from the drive’s entire prior lifetime rather than just its most recent use; a used drive showing early signs of degradation in these areas is a reasonable one to avoid or negotiate down in price rather than trust as a primary boot or game drive.
If a seller can’t or won’t provide a CrystalDiskInfo screenshot or similar SMART report before purchase, checking it immediately upon receiving the drive, before committing anything important to it, is a reasonable minimum precaution, and many marketplaces support return windows specifically intended to cover exactly this kind of post-purchase verification.
What causes SSDs to wear out faster than expected
Total bytes written (TBW) is the metric manufacturers use to rate SSD endurance, and it accumulates from every write operation, not just intentional file saves. Background processes like Windows Search indexing, antivirus scanning, and browser cache writes all contribute small amounts of write activity continuously, though on a well-configured modern system this background writing is a minor fraction of total wear compared to actual data usage.
Using an SSD as a page file or virtual memory swap location is normal and not a meaningful wear concern on modern drives; older guidance suggesting otherwise predates the significantly higher endurance ratings of current-generation NAND, so there is no need to disable or relocate the page file specifically to preserve drive life.
Running an SSD consistently very close to full capacity does increase write amplification (the controller needing extra internal read-modify-write cycles to find free space), which modestly accelerates wear compared to a drive with reasonable free space. Keeping at least 10-20% of a drive free is a sensible practice for both performance and longevity reasons, covered further in our storage capacity planning guide and general SSD buying guide.
Sustained heavy workloads, like using an SSD as a scratch disk for video editing or running a game server with constant logging, will consume endurance faster than typical gaming use, which is mostly reads (loading levels and assets) rather than writes. If your Percentage Used is climbing faster than expected for how the drive is used, checking what background processes are writing heavily is a reasonable next step.
Benchmarking alongside health checks
SMART health status and raw performance are related but separate things; a drive can show a perfectly healthy SMART status while still underperforming its rated speeds due to thermal throttling, a full drive slowing writes, or an outdated firmware version. Pairing a health check with a quick benchmark using CrystalDiskMark or AS SSD Benchmark gives a fuller picture.
CrystalDiskMark, from the same developer as CrystalDiskInfo, runs sequential and random read/write tests and reports results in MB/s or IOPS depending on the test preset chosen. Comparing results against the drive’s advertised specifications (available on the manufacturer’s product page) can reveal whether a drive is performing as expected or has degraded for reasons SMART data alone would not show, like thermal throttling under sustained transfers.
A meaningful gap between advertised and measured sequential speeds, more than roughly 15-20%, on a drive with clean SMART data is worth investigating separately; common causes include running in the wrong PCIe slot generation, a drive nearing full capacity (which slows write performance on many SSD controllers), or insufficient cooling on M.2 drives without a heatsink, covered further in our NVMe SSD guide.
When SMART data will not catch a problem
SMART monitoring is not infallible. It is fundamentally a self-reported diagnostic built into the drive’s own firmware, and while manufacturers generally implement it in good faith, it cannot predict every failure mode. Sudden controller failures, firmware bugs, or a drive simply disconnecting due to a bad cable or connector can all happen without any SMART attribute changing beforehand.
A drive that suddenly disappears from the system entirely, rather than showing degraded SMART attributes first, is a different problem than gradual wear; check physical connections, try a different SATA cable or M.2 slot, and confirm the drive shows up in the motherboard’s BIOS storage list before assuming the drive itself has failed outright.
Data corruption from a sudden power loss, particularly on drives without power-loss protection capacitors (a feature more common on enterprise drives than typical consumer gaming SSDs), can occur without triggering a SMART warning at all, since it is a data integrity event rather than a hardware degradation event. A reliable UPS for a gaming PC reduces this specific risk more effectively than any monitoring software can.
Troubleshooting SSD health check issues
If CrystalDiskInfo does not detect an NVMe drive at all, this is most commonly a driver issue rather than a hardware problem; ensure the drive is connected through a natively supported M.2 slot rather than through certain third-party PCIe adapters that some monitoring tools handle inconsistently, and confirm the drive appears in Windows Disk Management as a sanity check.
External USB SSDs sometimes show limited or no SMART data because many USB-to-SATA or USB-to-NVMe bridge chipsets do not pass SMART commands through transparently to the host software; this is a limitation of the specific enclosure or adapter rather than something CrystalDiskInfo can work around, and it does not necessarily indicate anything wrong with the drive itself.
A health status that suddenly drops from good to caution or bad after a firmware update is worth double-checking against the manufacturer’s own storage management utility if one exists, since in rare cases a firmware update changes how certain attributes are reported without an actual change in physical drive condition; cross-referencing two tools avoids a false alarm.
If a drive shows healthy SMART data but the system reports frequent read/write errors, corrupted files, or unexpected crashes tied to disk activity, consider that the issue may be a loose connection, a failing cable, or in rare cases a motherboard storage controller problem rather than the drive itself; swapping the same drive to a different port or cable is a useful isolation step.
Frequently asked questions
What health percentage should I worry about?
There’s no universal cutoff, but once CrystalDiskInfo’s overall status switches from blue (good) to yellow (caution), it’s worth backing up important data immediately even if the percentage still looks high, since the status reflects specific attribute thresholds, not a simple linear scale.
Does checking SMART data wear out an SSD?
No, reading SMART attributes is a passive query that does not write to the drive or consume write endurance; you can check as often as you like without any wear impact.
Is 100% drive health guaranteed to mean no problems?
A 100% or good status means SMART attributes are within normal range, but SMART cannot predict every failure mode, including some controller or firmware issues, so it’s a strong signal rather than an absolute guarantee.
Why does my new SSD already show less than 100% health?
Some manufacturers’ health algorithms start factoring in even minimal use immediately, or the drive shipped with firmware-level wear leveling already active from factory testing, so a 98-99% reading on a brand new drive is common and not a defect on its own.
Can CrystalDiskInfo damage my drive or void warranty?
No, it only reads data the drive’s own controller already tracks and reports through the standard SMART interface; it cannot modify firmware or void a manufacturer’s warranty through normal use.
Daniel Kovac, Components Editor, checks SMART data on every review drive before and after endurance testing on the PSU load tester bench, and the pattern holds consistently: attributes that stay flat for months and then start climbing are always worth acting on immediately rather than waiting to see if they level off. If you’re shopping for a replacement or additional drive, our guides to the best SSDs for gaming and best external SSDs cover current options.
Related guides
- Best Reviews of Piano in 2026: Specs and Prices Compared: 8 Tested (2026)
- Best Ge Profile Opal Mini Nugget Ice Maker in 2026: Top Picks for Every Budget: 8 Tested (2026)
- Best Chocolate Gift Box in 2026: Specs and Prices Compared: 8 Tested (2026)
- Best Microwave Convection Oven in 2026: Top Picks for Every Budget: 8 Tested (2026)
