Is My Mac’s SSD Failing? What SMART Can — and Can’t — Tell You

There’s a built-in health flag for your Mac’s drive, and there are real wear numbers underneath it — but both have honest limits, and the most dangerous myth is that a “Verified” status means you’re safe. Here’s how to read the real signals, what SMART genuinely tells you, and the one thing that actually protects your data.

By CoreGuard · Jul 7, 2026 · 14 min read

An SSD dying is the failure that actually costs you something — not time, but data. So it’s worth knowing how to read the warning signs, and, just as importantly, how much faith to put in them. The short answer is that your Mac has a genuine health signal for its drive, there are real wear numbers underneath it, and both are more limited than the internet usually admits. This is the honest version: what to actually look for, how to check it with built-in tools and one free utility, and why the reassuring green “Verified” is the most misunderstood word in the whole subject.

How do you tell if a Mac SSD is failing?

The clearest signals are a SMART status of “Failing” in Disk Utility, or reproducible trouble: files that won’t open or keep corrupting, files or folders disappearing, the drive suddenly going read-only, read/write errors, or the Mac failing to boot. Any of those means back up immediately. But here’s the honest catch that shapes everything below: an SSD can also fail suddenly, with no warning at all — so a current backup, not a health check, is what actually protects your data. A diagnostic tells you what it can see; it can’t promise there’s nothing it can’t.

The symptoms — and which mean “back up now”

SSDs are quieter than the spinning drives they replaced. There’s no click of death, often no gradual slowdown — sometimes the first sign is the failure. Still, these are the signals worth knowing, split by how urgently they call for a backup:

The read-only lockdown deserves a special mention: when an SSD’s controller runs out of healthy spare blocks, it often flips the whole drive to read-only on purpose — a last-ditch move to let you copy your data off before it goes entirely. If you can suddenly open files but can’t save, edit, or delete anything, don’t troubleshoot — copy your data to another disk right now. Symptoms like a lone kernel panic or a slow afternoon, by contrast, have plenty of innocent explanations (we covered the calm way to read those in “is my Mac dying?”), so correlate before you conclude.

Check 1: the SMART status in Disk Utility

Your Mac has a built-in, one-click health flag — you just have to click the right thing. Open Disk Utility (Applications › Utilities), choose View › Show All Devices, and — this is the part everyone misses — select the physical disk at the top of the tree (something like “APPLE SSD…”), not the “Macintosh HD” volume beneath it. The S.M.A.R.T. status line then appears in the info panel. Selecting the volume instead of the device is the number-one reason people say they “can’t find” SMART status.

You’ll see one of three words:

• Verified — the drive reports no fault. Reassuring, but read the next section before you exhale.
• Failing — the drive is predicting its own failure. Per Apple, the move is to “back up your data and replace the disk — you can’t repair it.” Do the backup first.
• Not Supported — the drive isn’t passing SMART to macOS at all. This is common for USB enclosures whose bridge chip doesn’t forward SMART; it’s not itself a bad sign.

The same status shows read-only under System Information › Storage (the exact menu path shifts a little between macOS versions). But note the ceiling here: Disk Utility’s SMART is pass/fail only. It gives you a green light or a red one and no wear data — no percentage used, no terabytes written, no spare-block count. As the macOS-internals writer Howard Oakley put it, in Disk Utility “all you’ll see is SMART status: Verified” — a single binary flag, not a health meter.

The trap: “Verified” is not a clean bill of health

This is the most important idea in the article, so I’ll be blunt: “Verified” means “no fault reported right now,” and nothing more. It is not a promise that the drive is healthy, and it is certainly not a reason to skip a backup.

The data backs this up. In Backblaze’s large hard-drive fleet, roughly 23% of drives that failed had shown no prior SMART warning at all — they were “Verified” until they weren’t. (That figure is from hard-drive data, not an SSD fleet — but SSDs have their own sudden failure modes too.) SSDs tend to fail suddenly — a controller fault, a firmware bug, a power event — and those failure modes can brick a drive instantly while its wear counters still read “98% healthy.” The reassuring flag and the dead drive are entirely compatible.

None of this means SMART is useless. A “Failing” status is a real, valuable warning — when the drive raises it, believe it. The mistake is reading the absence of a warning as a guarantee. Treat “Verified” the way you’d treat a smoke detector that isn’t beeping: good news, not a fireproof house.

Check 2: the real wear numbers, with smartctl

If Disk Utility only gives you a green light, where are the actual numbers — the wear, the terabytes written, the spare blocks? They exist; they’re just not in a built-in GUI. The standard way to read them is smartmontools, a reputable open-source utility, via its smartctl command. Install it with Homebrew and run it against the physical disk:

brew install smartmontools
diskutil list          # find the disk (the internal one is usually disk0)
sudo smartctl -a disk0

Here’s the part that a lot of older guides get wrong: on Apple Silicon, smartctl reads the internal SSD’s NVMe health log natively — no lowered security, no kernel extension. You get the endurance data Disk Utility hides. An abridged, illustrative readout looks like this:

SMART overall-health self-assessment test result: PASSED
Available Spare:                    100%
Available Spare Threshold:          10%
Percentage Used:                    4%
Data Units Written:     41,235,880 [21.1 TB]
Power On Hours:                     2,310
Media and Data Integrity Errors:    0

The honest caveat runs the other way from what you’d expect: it’s external USB drives that are the problem, not the internal one. Many USB-to-SATA and USB-to-NVMe enclosure bridges don’t pass SMART through on macOS, so smartctl may return nothing useful for an external drive even though it reads your internal SSD fine. Thunderbolt and NVMe enclosures generally do better. So: internal Apple-Silicon SSD, readable; a random USB caddy, often not.

What the wear numbers actually mean

Now the interpretation, which is where the no-verdict rule bites hardest. The key metrics, per NVM Express:

• Percentage Used — the drive’s own estimate of how much of its rated write life is spent. Think of it as a fuel gauge, not a clock. It can pass 100% and keep working; it’s an estimate, not a limit.
• Available Spare (vs its threshold) — the reserve NAND blocks still in hand. A steadily falling spare, especially near the threshold, is a genuine wear signal.
• Data Units Written — total lifetime writes, i.e. your terabytes written (TBW). Compare it to the drive’s rated endurance for a sense of how hard it’s been used.
• Media and Data Integrity Errors — should be 0. Any non-zero value is worth attention.

Crucially, none of this is a countdown. The macOS-internals writer Howard Oakley, who has measured this carefully, frames SSD life as expectancy, not a death date — in one of his worked examples a normally-used internal SSD had enough write headroom left to last well over a century. For the vast majority of people, write endurance is nowhere near the thing that will retire the Mac. We went deep on wear, TBW, and why free space isn’t the number to watch in “free space is a lie,” and on how a drive’s identity can be faked in “the label is not the drive.” The catch, again: a low Percentage Used doesn’t protect you from a sudden controller or power failure. Wear is a trend to watch, not a promise.

First Aid checks the file system, not the hardware

One more built-in tool, and one more honest boundary. Disk Utility’s First Aid is often reached for when a drive acts up, but it’s important to know what it does. In Apple’s own words, First Aid “can check for and fix errors related to the formatting and directory structure” of a drive — the file system, a logical layer — and Apple adds that it “can’t detect or repair all problems.” It does not measure NAND wear, and it does not repair or predict physical failure.

So First Aid and SMART answer two different questions. A drive can pass First Aid and still be failing (the file system is tidy; the hardware is dying), and it can show file-system errors while the hardware is perfectly fine. Run First Aid bottom-up — each volume, then the container, then the device — for logical errors; just don’t read a clean pass as proof the hardware is healthy.

Back up first — and back up now if it’s misbehaving

Everything above points at one conclusion: the check is not the protection, the backup is. Keep a running backup — Time Machine to an external drive, or a clone — on the old 3-2-1 principle (three copies, two kinds of media, one off-site), and don’t let a “Verified” status talk you out of it.

And there’s one situation that overrides every other step in this article: if the drive is already misbehaving — SMART says “Failing,” it’s gone read-only, you’re seeing read/write errors, or files are vanishing — back up now, before you run a single diagnostic. A dying SSD gets worse with every write, and First Aid, a reformat, or a “cleaner” all write to the drive. Stop non-essential use, minimize writes, and copy your irreplaceable data to another physical disk first. The worst possible order of operations is “troubleshoot first, discover the drive was dying second.”

Repair or replace?

If a drive really has failed, the honest framing is arithmetic, not a verdict — and one principle sits above it: your data is worth more than the Mac. If you backed up (you did, right?), the hardware is just a replaceable object.

The mechanics depend on the drive. An external or older upgradeable SSD is a normal swap — replace the drive. But the internal SSD on an Apple-Silicon Mac is soldered to the logic board and cryptographically paired to the Secure Enclave; there’s no module to change. A genuinely failed internal SSD there means a logic-board (or whole-unit) replacement — which is precisely why the backup, not the repair, is what saves you. Specialist data-recovery shops can sometimes rescue data from a failed drive, but that’s an expensive last resort that works best when you stopped writing to the drive early.

What not to do

Searching “is my SSD failing” drops you into a field of bad advice. Skip all of it:

• Don’t trust a “health score,” a failure date, or a “dies in N days” countdown. SMART can’t support that; it’s false precision dressed up as certainty.
• Don’t run a “cleaner” or “optimizer” to “fix” a failing drive. You can’t repair failing hardware in software, and the extra writes make a dying SSD worse.
• Don’t delete files or caches hoping to “heal” it. That’s writing to a drive you should be reading off of.
• Don’t read “Verified” as “safe forever” — and don’t assume passing First Aid cleared the hardware.
• Don’t keep hammering a suspect drive with benchmarks or a reformat before your data is copied somewhere else.

Where CoreGuard fits

This whole article is the thinking behind CoreGuard, so let me be plain about what it does. The honest readings here are scattered: a pass/fail flag in Disk Utility, the real wear numbers only via a Terminal command most people won’t run, First Aid in a third place. CoreGuard reads the drive’s own data — the SMART Verified/Failing status and, where the drive exposes it (including the internal Apple-Silicon SSD’s NVMe endurance log), the wear underneath — and puts it in one place, in plain English. “Verified” stops being a black box. The basic health verdict and every danger or health warning are free, forever — including that a drive is worn or failing; the detailed endurance readout (terabytes written, power-on hours, raw SMART, spare, the wear trend over time, and a “years-left” estimate) and the exportable report are Pro.

And the hard line, because it’s the reason this exists: CoreGuard never invents a health score, never predicts a failure date, and never claims to see wear the platform genuinely hides — when a metric isn’t available on your Mac (say, a USB enclosure that blocks SMART), it says so rather than guessing. It won’t “clean,” “optimize,” speed up, or repair your drive, and it isn’t an antivirus. It shows you the drive’s numbers, explains what they mean and what they don’t, and reminds you that the backup is the protection. It’s local-only: zero network connections, verifiable with lsof -i -nP | grep CoreGuard.

CoreGuard isn’t out yet — the download and checkout go live shortly. Free covers the live SMART status, the basic SSD-health verdict, and every danger warning, forever; Pro adds the detailed endurance numbers — terabytes written, power-on hours, raw SMART, and the wear trend — plus the exportable report, and is a one-time $29 (Family $49), perpetual, not a subscription, with a 30-day money-back guarantee. If you want your drive’s real numbers in one honest place, get notified and grab it free at launch, or see what Pro adds. Either way — check the status, read the wear as a trend, and keep the backup running.

Frequently asked questions