How to Fix a Corrupted USB Drive | Three Scenarios and What to Try First

Table Of Contents show

A USB flash drive that shows up as “corrupted” is telling you something is wrong — but corruption is a symptom, not a root cause. Three very different underlying problems produce the same symptom on your computer screen, and the right response to each is different. Software recovery works well for one of the three. The other two require lab-level hardware work, and running software against them tends to make the problem worse. Before you download anything or try to reformat the drive, it’s worth two minutes to figure out which scenario you actually have.

The three scenarios that produce a “corrupted” USB drive are:

Interrupted write from unsafe removal — the file system is genuinely corrupt on an otherwise-healthy drive. This is the most common by a wide margin, and software recovery works well.
NAND-level “soft” corruption from failing flash memory — the file system looks corrupt because the underlying storage is delivering bad reads. Software recovery tends to make this worse. Hardware work first, recovery second.
Component failure that makes the volume appear corrupt — the drive is misreporting its own capacity or geometry because of a hardware fault, and the file system on it looks broken as a result. Fix the hardware and the “corruption” often disappears.

The rest of this article covers how to tell which one you have and what to do about each. If you already know your drive has physical damage, the file system is only the tip of the problem — skip ahead to the hardware-symptom guidance below or go directly to our USB flash drive recovery pillar.

Start a free hardware recovery evaluation →

What “corrupted” actually means

Every USB flash drive has two layers that matter for corruption. Underneath is the raw NAND flash memory, where your data physically lives as electrical charges in memory cells. On top of that is a file system — usually FAT32 or exFAT for USB drives — which is a set of metadata tables that tell the operating system where each file starts, how big it is, and which memory blocks store its contents.

When Windows or macOS tells you the drive is corrupted, what it’s reporting is that the file system on top doesn’t match what the storage layer underneath is delivering. There are only a few ways for that mismatch to happen. Either the file system itself is broken (Scenario 1), or the storage is delivering unreliable reads that the file system parser can’t make sense of (Scenario 2), or the drive is reporting the wrong storage layout entirely (Scenario 3).

Which one you have determines whether software can help.

Scenario 1: Interrupted write from unsafe removal

This is the most common cause of a corrupted USB drive by a large margin, and it’s the one where software recovery genuinely works.

FAT32 and exFAT — the file systems used on almost every consumer USB flash drive out of the box — have no journaling. Journaling is a database technique where a file system writes down what it’s about to do before it does it, so that if the operation is interrupted, the file system can either finish the operation on the next mount or roll it back cleanly. Modern desktop file systems like NTFS, ext4, APFS, and Btrfs all journal. FAT32 and exFAT do not.

What this means in practice: if you pull a USB drive out of the port while Windows was still writing to it — even if the copy dialog looked finished a moment earlier, because Windows caches writes and finishes them asynchronously — the file allocation table on the drive can end up in a partial state. Some files might be half-written. The directory table might reference files whose actual data blocks are marked free. Cross-linked file entries can point at the same physical block. All of these show up to the operating system as “the file system is corrupted” and produce error messages like “you need to format the disk in drive X: before you can use it” or “the file or directory is corrupted and unreadable.”

The underlying storage — the actual NAND memory and the controller and the USB connector — is fine. It’s only the file system metadata layer that’s been left in an inconsistent state.

Software recovery is the right tool for this. Recovery software reads the raw sectors of the drive, looks for file signatures independent of the broken file system, and reconstructs a coherent view of what was on the drive. Most files come back intact because the actual file data blocks were written successfully — only the pointers to them were broken.

Where to try software recovery: If you’re not comfortable running data recovery software yourself, the fastest path is walking the drive into a local computer repair shop. Any competent local shop can run through the standard logical recovery workflow on a USB drive that’s otherwise healthy. If you’re in a metro area with a Micro Center store, their in-store data recovery service is a partner of ours and handles this kind of logical corruption case reliably and quickly. Most cases resolve for a fraction of what a professional-lab recovery would cost, and turnaround is same-day or next-day rather than the multi-day timeline of a lab case.

If local software recovery fails — either the shop tells you the drive isn’t recoverable, or the recovered files are all corrupted when you try to open them, or something about the case turns out to be more complicated than pure logical corruption — the drive is likely dealing with one of the other two scenarios below, and Gillware can still handle it. Cases that fail at the software layer sometimes reveal underlying hardware issues that only professional lab equipment can address.

Scenario 2: NAND-level “soft” corruption from failing flash memory

The second scenario looks similar to the first — the drive mounts, the file system reports errors, files come back corrupted — but the root cause is fundamentally different. Instead of the file system layer being broken on top of healthy storage, the storage itself is delivering unreliable reads.

NAND flash memory stores data as electrical charges in individual memory cells. Over time, and especially in older drives or drives used at high temperatures, those cells can lose their ability to hold charge reliably. A cell that was written as a 1 slowly drifts toward 0 as it leaks charge. The drive’s controller normally corrects this with error-correction coding, but past a certain threshold of degradation, the error correction fails on some reads and the controller returns marginal data to the computer.

The computer sees this as file corruption — individual files fail to open, or open with garbled content, or the file system itself throws errors during scans because the metadata blocks are being delivered inconsistently.

Why software recovery makes this scenario worse: Every read of a degraded NAND cell degrades it further. Recovery software that repeatedly scans a drive with failing NAND is putting stress on the exact cells that are already marginal. On top of that, some recovery software will attempt to write “fixed” file system metadata back to the drive as part of its repair process — and writes to a failing NAND chip can destroy blocks that would otherwise still be readable.

The Gillware approach: When we identify a drive as being in NAND-level soft corruption territory, the first priority is to stop the ongoing degradation by imaging the drive at the raw memory level under controlled conditions, not through the normal USB interface. Sometimes we can address the underlying hardware issues — controller settings, voltage tuning, error correction thresholds — and get clean reads off the controller that then produce a usable file system without further intervention. When the NAND itself has degraded past the point where the controller can deliver clean reads, we move to chip-off recovery, extracting the NAND chips directly and reading them with voltage-tuning strategies designed for degraded flash.

Scenario 3: Component failure making the volume appear corrupt

The third scenario is the one customers rarely diagnose correctly on their own. The drive plugs in and enumerates — the computer sees it and assigns a drive letter — but the reported capacity is wrong. A 64 GB drive might show as 32 GB, or 8 GB, or 500 MB. The volume shown at the wrong capacity looks corrupt to the operating system because the file system stored on the drive references block addresses beyond what the drive is now reporting as available. Directory entries point at data blocks that appear not to exist. Windows offers to format the drive.

The corruption is real from the operating system’s perspective, but it’s a downstream symptom, not the actual problem. The actual problem is a hardware fault — typically in the controller chip or in the NAND-to-controller signaling — that’s causing the drive to misreport its own geometry. The data on the drive is fine. It’s just that the drive can’t currently tell the computer where to find it.

Why software recovery doesn’t help this scenario: No amount of file system repair addresses the misreported capacity, because the capacity is being decided by the drive itself before any file system operations happen. Recovery software might extract some files from within the wrongly-reported range but will fail to extract anything from beyond it. Format prompts are dangerous here — formatting to the wrong capacity locks in the incorrect geometry and makes recovery meaningfully harder later.

The Gillware approach: When our engineers identify a Scenario 3 drive at intake — the signature is the mismatch between the drive’s labeled capacity and its reported capacity, combined with an otherwise-functional connection to the computer — we address the controller-side hardware issue first. Once the drive is reporting its correct geometry again, the file system that appeared corrupt reads as a completely normal file system. What looked like a lost drive with corrupted data is often, after the hardware fix, a fully-mountable volume with intact files.

How to tell which scenario you have

You can narrow down which of the three scenarios applies without opening the drive or running any software.

Check the reported capacity first. Right-click the drive in File Explorer, look at Properties, and compare the reported size to the labeled capacity on the drive (or on the packaging or in your purchase records). If the drive is a labeled 64 GB drive and Properties shows “about 60 GB” (accounting for the standard formatting overhead), you’re looking at Scenario 1 or 2. If it shows a dramatically wrong size — 8 GB when it should be 64 GB, or 0 bytes, or nothing at all — you’re looking at Scenario 3 or a more severe hardware failure.

Try the drive on a different computer. If the same drive corrupts on multiple computers in the same way, the issue is with the drive rather than with any particular computer’s operating system or USB stack. If the drive works fine on a different computer, the problem might be on the original computer’s side rather than the drive’s.

Look for physical damage. A visibly bent, cracked, or snapped USB connector, a housing that’s come apart, discoloration or char marks from heat exposure, or evidence of water damage all point away from pure logical corruption and toward hardware issues that require lab work. Physical damage is often accompanied by file system errors as a secondary symptom, but the hardware needs to be addressed first.

Check when the drive was purchased. A drive less than a couple of years old with typical light use is unlikely to be in Scenario 2 (NAND wear-out). A drive that’s been in continuous use for five years or more, or a drive kept in a hot environment, is much more likely to be showing NAND-level degradation. Age isn’t definitive but it shifts the odds.

Consider what you were doing when it broke. If the corruption showed up right after you pulled the drive out mid-copy or the computer crashed while writing to the drive, Scenario 1 is likely. If the corruption showed up out of nowhere — the drive worked yesterday, doesn’t today, no incident in between — hardware failure is more likely.

What to do first if you’re in Scenario 1

Assuming your drive shows the correct capacity, mounts and enumerates normally, and just has file system errors that appeared after an unsafe removal or write interruption:

Stop writing to the drive. Any additional writes can overwrite the data blocks that hold your files. Don’t run repair prompts (“scan and fix,” “format the drive to use it”) — those are writes.
Take the drive to a local computer repair shop or a Micro Center. A logical recovery on a healthy drive is a routine job that most shops can complete same-day. Micro Center’s in-store data recovery service handles logical corruption cases reliably. The cost is meaningfully less than a professional lab recovery and the turnaround is faster.
If the shop can’t recover it, come to Gillware. Sometimes what looks like Scenario 1 turns out to have Scenario 2 or 3 underneath. If a local shop’s software recovery fails or produces corrupted files, that’s a signal that the underlying hardware needs professional attention. We can pick up cases that started as software recovery attempts and diagnose whether the deeper issue is hardware.

When to skip software and come straight to Gillware

Some situations don’t belong in the software-recovery workflow at all. Send the drive directly to a professional lab if:

The drive has visible physical damage. Snapped connector, cracked housing, char marks, moisture inside, evidence of impact. Software can’t address physical damage.
The drive doesn’t enumerate at all. Windows shows “USB Device Not Recognized,” Device Manager surfaces a Code 43, or the drive isn’t detected. This is hardware territory — usually controller failure — and software recovery has no entry point.
The drive shows dramatically wrong capacity. A 64 GB drive reporting 8 GB, or 0 bytes, or nothing at all — this is Scenario 3 or worse, and software recovery will only extract from the wrongly-reported range.
The drive was through fire, water, extended crushing, or a static discharge event. The hardware needs to be assessed and addressed before any data can be pulled.
A prior repair attempt failed. If a phone repair shop or a DIY attempt has already been made and the drive is now in worse condition, the repair-side path is closed and recovery has to bypass the damage entirely.
The data is irreplaceable or high-value. A single attempt with the wrong tool can make recovery impossible. When the data value warrants professional treatment, skip the software step.

What Gillware does for hardware-caused corruption

When a drive arrives at our lab with the corruption symptom and one of the hardware-cause signatures, our workflow is:

Free intake diagnostic. The drive is examined under stereo microscope, tested on isolated diagnostic hardware, and evaluated for which of the three scenarios (or which combination) applies. This step is always free.
Firm written quote. After diagnosis, we quote you a firm price for the recovery path that applies — no ranges, no estimates that grow as the work progresses. You decide whether to proceed.
Recovery work. Depending on scenario, this might involve micro-soldering if the connector-side is involved, controller-level intervention for capacity misreporting, or chip-off recovery for NAND-level failure. Our flash drive recovery pillar explains the paths in detail.
Data return. Recovered files are returned on new media or transferred securely. The original drive is not returned to service.

Standard cases operate under our “no data, no charge” engagement: if the recovery is unsuccessful, you don’t pay for the work. Cases involving significant additional engineering are quoted individually before work starts.

Start your hardware-caused corruption recovery

If your USB drive is showing corruption symptoms that point to hardware — wrong reported capacity, physical damage, prior repair failure, or software recovery that came up empty — the next step is a free evaluation.