Yes, files on an external drive sitting in a drawer can silently corrupt, and your annual S.M.A.R.T. check will not detect it. Tools like CrystalDiskInfo report the condition of the drive's mechanics and electronics, not the integrity of your files. Detecting bit rot requires comparing your data against a known-good reference, such as checksums or a verified copy, and preventing it requires multiple independently stored copies that you verify and refresh on a schedule.
What bit rot actually is
"Bit rot" is the gradual loss of data that was previously readable, without any hardware failure event. All storage media fade over time: the magnetic domains on hard drive platters and tape weaken, the electrical charges trapped in SSD cells leak away when the drive is unpowered, and the organic dyes in recordable CD and DVD media degrade. There is no practical way to make any of these media permanent.
It helps to separate two related problems. A drive-level failure means sectors become unreadable or the mechanism fails; S.M.A.R.T. attributes such as reallocated and pending sector counts usually reflect this. A file-level corruption means the bytes come back wrong even though the drive believes it read the sector successfully. Both destroy data, but only the second one happens with no warning from the drive at all.
Why your annual S.M.A.R.T. check misses it
CrystalDiskInfo and similar tools read attributes that predict mechanical and electronic failure: reallocated sectors, read errors, temperature, spin-up time. These are worth checking, because a drive with a rising pending-sector count is telling you it is dying. But a flipped bit inside a sector that still reads "successfully" produces no attribute change whatsoever. A drive can pass every S.M.A.R.T. check with a clean bill of health while individual files rot.
A full surface scan with a tool like HD Tune reads every sector and will find sectors that have become unreadable or slow, which is genuinely useful as an early-warning step. What it cannot do is tell you whether the data coming back is the data you originally wrote. Only a comparison against an independent reference can do that.
Detecting bit rot with checksums
The practical detection method is to compute a cryptographic hash of every file, store that list on a different device, and re-verify it each year. Any mismatch means that file changed without your action, so you restore it from another copy. On Linux or macOS:
cd /mnt/external
find . -type f -print0 | sort -z | xargs -0 sha256sum > ~/archive-manifest.sha256
# one year later:
sha256sum -c ~/archive-manifest.sha256On Windows, PowerShell's Get-FileHash -Algorithm SHA256 computes the same values per file.
A second approach is to take a complete image of the drive and compare the image against the source. General-purpose comparison tools such as WinMerge can diff every file on the drive against a mounted image, and forensic utilities go further, flagging differences even in unallocated space. If you keep several images made at different times, you can resolve a disagreement between the source and one image by majority vote, the same principle used for redundant computers on spacecraft. Checksum manifests are usually easier to automate, but imaging has the advantage of giving you a restorable snapshot in the same step.
Preventing bit rot: redundancy, environment, refresh
Since no medium lasts forever, prevention is really about making sure no single copy's decay can destroy your data:
- Keep at least two or three copies on different devices or media, with at least one stored elsewhere. Cloud storage can serve as one of these legs, but do not rely on it alone; providers have suffered outages and, occasionally, outright data loss.
- Control the environment. Store drives cool, dry, and stable, away from chemicals, mold, strong electromagnetic fields, and radiation.
- For actively used data, use a RAID array with scheduled scrubs. Periodic checks rewrite data that has become marginal before it fails entirely and alert you when a disk dies outright. Remember that RAID is not a backup; pair it with offsite copies.
- Consider tape for large long-term archives. Tapes minimize the complex moving parts that can fail, and rotating several tapes means a bad recent backup still leaves you an older good one. Verify readability and checksums after every write.
- Refresh aging copies. Copying data onto fresh media periodically resets the clock, because verified media still ages.
A realistic annual routine for a drawer-stored drive
- Connect the drive and record its S.M.A.R.T. values, watching for trends in reallocated and pending sectors.
- Run your checksum verification against the stored manifest.
- Restore any mismatched files from another copy, then investigate why they diverged.
- Optionally run a full read scan to catch weak sectors early.
- If files legitimately changed, regenerate the manifest and refresh at least one other copy.
- Disconnect and store the drive; an annual power-up is sufficient, and there is no need to leave it running.
Frequently Asked Questions
Can bit rot affect a drive that is never used?
Yes. Magnetic decay depends on time and storage conditions, not on how often you read the drive. Light use does not harm an archival drive; neglecting verification does.
Does copying a file repair bit rot?
Copying to a different drive gives you a fresh, independently written copy, which is the point. Copying within the same drive does not restore degraded magnetic signal, and it cannot fix bytes that were already wrong.
Where should I keep the checksum manifest?
On separate media from the data, ideally in more than one place. If the manifest rots alongside the archive, you lose your reference for detecting corruption.
Are SSDs safe for long-term drawer storage?
No, they are riskier. Unpowered SSDs leak cell charge and consumer drives are typically only specified to retain data for around a year at room temperature. Rewrite SSD archives regularly or prefer magnetic media for cold storage.
How often should I verify?
Once a year is a reasonable minimum for cold storage. Verify more often for irreplaceable data, large collections, or drives kept in warm or humid conditions.