Updated on Oct. 15, 2026
Position: Resource - Disk Utilities - How to Test USB Flash Drive Speed on Windows 11/10?
Your USB flash drive says USB 3.0 on the box, with "up to 400 MB/s" in small print underneath. Windows copies a folder at 28 MB/s and the progress bar barely moves.
Both numbers can be correct. The advertised speed comes from a bench test under ideal conditions. What Windows reports is what happened on your machine, through that port, with those files. Only the second one tells you what the drive will actually do for you.
The fastest way to find out is DiskGenius. Plug the drive into a USB 3.0 port, open DiskGenius, select the drive in the left panel, choose "Tools" → "Disk Speed Test" and start. You get sequential and random read and write speeds, IOPS, access time, and a speed curve for each pass, in a few seconds.
Table of Contents:
Most people who run a benchmark and feel let down were never going to see the number on the box. That number was describing the interface, not the drive. Knowing what you are actually aiming at saves you from that disappointment.
| Interface | Theoretical Max | Typical Real-World Read | Typical Real-World Write |
| USB 2.0 | 480 Mbps (60 MB/s) | 20 to 35 MB/s | 10 to 20 MB/s |
| USB 3.0 / 3.2 Gen 1 | 5 Gbps | 80 to 160 MB/s | 30 to 100 MB/s |
| USB 3.1 / 3.2 Gen 2 | 10 Gbps | 200 to 450 MB/s | 150 to 350 MB/s |
| USB4 / Thunderbolt 3-4 | 40 Gbps | 1,000 to 3,000 MB/s | 800 to 2,500 MB/s |
The theoretical number is the speed of the road, not the car. A budget USB 3.0 stick in a USB 3.0 port can still write at 15 MB/s because the flash memory inside is slow. Everything from the port to the NAND chips has to be fast, and the slowest link sets the pace.
USB 3.x on the spec sheet describes the connector and the bus. It says nothing about the flash, the controller or the cache inside the drive. A USB 3.2 sticker on a $6 stick is not a speed promise.
These ranges come from drives are not from any official standard. Use them as a sanity check.
| Drive Tier | Sequential Read | 5 to 15 MB/s |
| Budget USB 2.0 stick | 15 to 30 MB/s | 5 to 15 MB/s |
| Budget USB 3.x stick | 40 to 100 MB/s | 15 to 40 MB/s |
| Mid-range USB 3.x | 100 to 250 MB/s | 40 to 120 MB/s |
| High-end USB 3.x | 250 to 500 MB/s | 150 to 400 MB/s |
| NVMe-based USB drive | 800 to 2,000 MB/s | 500 to 1,500 MB/s |
Read is almost always faster than write, often by 30 to 50 percent, and manufacturers quote the read number because it is bigger. A drive advertised at 400 MB/s might write at half that even in the best case. Capacity plays a role too. The 256 GB version of a drive family is usually faster than the 32 GB version because larger drives have more flash chips working in parallel.
Large video files and photo shoots punish slow write speeds because you sit and wait for the copy. A backup that takes 8 minutes on a fast drive takes 45 on a slow one, and people start skipping backups when they hurt.
Copying a folder of thousands of small documents is a different bottleneck entirely. It is dominated by per-file overhead and random access rather than throughput. A drive that looks fine on one large video can crawl through a folder of text files.
Booting a Windows installer or running portable apps from the drive leans on random read and access time. A drive that scores well on sequential tests can still feel sluggish for this, and it is worth checking the random numbers before you decide which drive to buy.
A speed test also works as a health check. If a drive that used to copy at 90 MB/s now does 12, something changed. Better to find out while it still works than after it starts corrupting files.
Megabytes per second and megabits per second are different units. 8 bits make a byte, so a USB 2.0 link rated at 480 Mbps is a 60 MB/s link before overhead, and in practice tops out near 35 MB/s.
Product pages often use Mbps because the number is eight times larger. When a spec sheet says 480 Mbps and your test reports 35 MB/s, you are seeing the same thing. Divide by 8 when the unit is Mbps, and check which unit your benchmark tool reports. They do not all agree.
Plug the drive straight into the computer, ideally a rear port on a desktop rather than a hub or monitor passthrough. A cheap hub or a long extension cable will throttle the link, and the test will measure the hub instead of the flash drive.
Then confirm the port is really USB 3.0. Blue plastic inside the port is the usual hint but not a rule. In Device Manager, expand "Universal Serial Bus controllers" and look for entries with "USB 3.0" or "xHCI" in the name. A USB 2.0 controller will show "Enhanced Host Controller" or "USB 2.0" instead. If the drive appears under a USB 2.0 hub node, that is your bottleneck.
We have also run into drives that should be USB 3.0 but negotiated USB 2.0 on first plug-in. Unplugging and reinserting usually fixed it. If the port is fine but the speed is capped near 35 MB/s, that is worth trying before anything else.
Anything touching the disk eats into the numbers. Antivirus real-time scanning inspects every file written to the drive. Cloud sync tools may read the folder you are copying from. Windows Update, a Steam download or a video render will do the same to the source disk.
Give the machine a quiet minute before starting, and do not use it heavily during the test.
Read tests are safe. Write tests are not always safe, and some tools include destructive modes that write patterns across the whole device.
Copy anything irreplaceable off the drive first. Then read the labels carefully before clicking start, and confirm the target by drive letter, size and model name rather than by position in a list.
DiskGenius is a Windows disk utility that includes disk benchmark alongside disk partitioning, backup and data recovery features. The practical benefit is that you can measure the drive now and keep working with the same program if the results suggest something worth investigating.
Step 1. Download DiskGenius from the official site and install. It supports Windows 11, 10, 8 and 7.
Step 2. Connect the USB drive and confirm in the left panel that you have the right one. The panel lists internal disks alongside removable drives. Check the model name and capacity rather than trusting the drive letter.
Step 3. Select the drive and open "Disk Speed Test" from the "Tools" menu.
Step 4. Choose "File Test" or "Benchmark Test" based on your needs, then click "Start Test".
The defaults are fine for a first look. Larger data sizes reveal the sustained write speed by pushing past the SLC cache. Smaller block sizes show how the drive handles small operations.
Step 5. Start and wait. Results cover sequential read and write, random read and write, IOPS and access time, with a speed curve for each pass.
Reading the output: sequential numbers map to copying a large file, random 4K numbers map to working with many small ones, and access time tells you how fast the drive responds to a request. If the write curve starts high and drops sharply partway through, the SLC cache filled up and the flat part of the curve is the real sustained speed.
DiskGenius also shows the partition layout, file system and capacity details from the same window. That comes in handy when a drive behaves oddly, because sometimes the problem is not hardware at all. A drive with file system errors, an unexpected partition structure, or a capacity that does not match the label will test poorly or act erratically.
Keep in mind that a benchmark measures performance, not health. A fast result does not prove a drive is in good shape and a slow one does not prove it is dying. It is one piece of evidence.
Read mode: How to Check USB Drive Health: 7 Ways to Test a Flash Drive?
Windows will give you a rough number with nothing but a stopwatch.
Step 1. Find or make a file of at least 1 GB. A single video file works well.
Step 2. Copy it from the computer to the drive. That is the write speed.
Step 3. Copy it back. That is the read speed.
Step 4. Divide the file size in MB by the seconds it took. 2,000 MB in 40 seconds is 50 MB/s.
Step 5. Repeat with a folder of several thousand small files if that is how you actually use the drive.
The Windows copy dialog shows a live transfer speed while it runs, which is enough for a rough check. If you want a cleaner number, time the whole transfer yourself, because the dialog speed is an instantaneous reading that jumps around.
This approach is approximate. Windows caches writes, so short transfers can look faster than what the flash is really doing. The source disk also has to keep up, and background activity moves the number. But it reflects real file transfers, which is what most people care about.
CrystalDiskMark is another widely used benchmark tool on Windows. You can download it from the official site, pick the USB drive from the dropdown, set the number of test runs and the test data size, and click the button. First row is sequential. Following rows are random 4K at different queue depths and thread counts.
For USB drives, set the test size to at least 1 GiB. That default is right on the edge of many drives' SLC cache, so a larger test like 4 or 8 GiB shows sustained speeds instead of burst. Queue depth of 1 gives the everyday picture. Higher queue depths represent workloads that flash drives rarely see.
USBDeview from NirSoft lists every USB device the computer has ever seen, with vendor IDs, serial numbers, connection history and speed information. Some builds include a read and write speed test.
Download it from NirSoft's own site, launch, and find the flash drive in the list. If the speed test function is present in your version, select the device and run it. It writes a temporary file to the drive and reports read and write speeds.
What this tool is really good for is confirming the connection. The device list shows whether a drive is linked at high speed (USB 2.0) or super speed (USB 3.x), which answers the "why is this slow" question in about five seconds. The speed test itself is basic and reflects one file under one set of conditions, so treat it as a quick check.
Older utilities like Check Flash (ChkFlsh) combine a speed test with read and write verification. Some versions can stress test a device until it fails, which people used to do to burn in new drives.
These are older projects. Download sources vary in trustworthiness, and compatibility with current Windows versions is not guaranteed. If you already use one and it works, fine. If you are starting fresh, Method 1 or Method 3 covers the same ground with current software and clear documentation.
If you do run a legacy tool, check where the file came from and prefer the original site. Read what each mode does before clicking, because some write test modes overwrite the entire device. And if a tool reports errors, look at the wider picture before condemning the drive. One failed pass on a warm, full, cheaply built stick is not proof of hardware failure.
Your benchmark tool will output several numbers. Here is what each one is telling you.
Reading is easier for flash memory than writing, so read speed is nearly always higher. Which one matters depends on what you do with the drive. If you hand off files to someone else, read speed decides how long they wait. If you copy files onto the drive, write speed decides your wait, and it is usually the number to watch.
A 30 to 50 percent gap between read and write is normal. A gap of 5x or more usually means the drive has a small write cache and little else, which is common on budget models.
Sequential speed is what you get moving one large file. Random 4K speed is what you get touching small pieces of data scattered across the drive, which is what happens when you copy a folder of photos, run an app or boot an installer.
IOPS, input/output operations per second, is the same random workload expressed as a count. Higher is better. Modern internal SSDs reach hundreds of thousands. Typical flash drives land in the low thousands.
Access time is the delay before the first byte arrives, in milliseconds. It looks tiny but multiplies across thousands of operations. A drive with strong sequential numbers and a poor access time will feel fast on a movie and sluggish on a folder of documents.
When comparing two benchmarks, match the workloads. A sequential number from one tool is not comparable with a random number from another.
Many flash drives start writing quickly and then slow down. The first few gigabytes go into a fast SLC cache, and once that fills, data goes directly to slower memory. If you have watched a copy start at 150 MB/s and settle at 35, that is what happened.
To find the sustained speed, use a data size larger than the cache. A 10 GB test file on a drive with a 6 GB cache shows both phases in one run. Filling the drive past half full and retesting works too, since the cache shrinks as free space runs out. Sustained write speed is the number that matters for long backups and video recording.
The advertised figure is often peak read speed under ideal conditions, sometimes measured on a bench rig rather than in a computer. The port or an adapter may be throttling the link. The cache might be exhausted and the drive has dropped to its base speed. Or the test conditions simply differ: file size, file system, remaining free space, temperature and background activity all move the number.
If your test is within 20 to 30 percent of the advertised read speed, the drive is doing what it said. The last few percent are usually about the test setup rather than the hardware.
Speed (MB/s) = File size (MB) ÷ Transfer time (s)
A 2,000 MB file in 40 seconds is 2,000 ÷ 40 = 50 MB/s.
If the spec sheet uses Mbps, divide by 8. Some tools report MiB (binary megabytes) instead of MB (decimal megabytes), which is about a 5 percent difference. Small enough to ignore for a sanity check, large enough to matter when you are comparing two tools side by side.
The Drive Is Connected to a Slower USB Port
The most common cause. A USB 3.0 drive in a USB 2.0 port tops out near 35 MB/s regardless of how good the drive is. Hubs, extension cables, front panel ports on older cases and USB-C to USB-A adapters can all downgrade the link. Try a different port, ideally a rear one, and check the connection speed in Device Manager.
The Drive Has Slow Write Performance by Design
Some drives are built for capacity and price rather than speed. Single channel controllers, low grade NAND, no cache. They read acceptably and write poorly. Nothing is broken and no software will fix it. If the spec sheet lists only a read speed, that is a hint.
The Drive Is Almost Full or Has Too Many Small Files
Flash drives slow down as they fill. The controller has fewer free blocks to work with, wear levelling gets harder, and the SLC cache shrinks. Past 80 or 90 percent full, expect a noticeable drop.
Small files add per-file overhead: opening, metadata, closing. Copying 4,000 files of 1 MB each is much slower than copying one 4 GB file even though the total is the same. Zipping the folder first often solves it outright.
The Drive Has File System Errors or Bad Sectors
File system corruption and bad sectors/blocks show up as slow transfers, files that refuse to copy, or a drive that disconnects mid-task. There is a difference between "this drive is slow" and "this drive is failing", and the second is worth catching early.
If a drive reports read errors, shows corrupted files or repeatedly disappears from Explorer, stop doing write tests and copy the important data off first. Repeated writes to a failing device make things worse.
Once the data is safe, you can check the condition. DiskGenius has a Check or Repair Bad Sectors function that scans the device and shows results in a grid, along with the drive's disk and partition information. That gives you something concrete to judge on.
Repairing bad sectors on a flash drive is not a reliable fix, since flash memory does not recover the way a hard disk sector sometimes can. A clean scan is not a guarantee either. Weigh it with the symptoms and the age of the drive.
The Drive May Be a Fake Capacity Counterfeit
Cheap drives sold online sometimes report more capacity than they actually contain. A 256 GB stick with 32 GB of flash will appear to work until you fill it past 32 GB, at which point older files silently corrupt. Slow speeds, write errors, and files that fail to open after copying are common symptoms.
Copy more data to the drive than the claimed capacity should allow, then open the files. Anything broken means the drive is lying about its size.
Read more: Why Is My SSD So Slow? 15 Causes and How to Fix Each One?
If the drive is slower than it should be, try these before replacing it.
Try a different USB port first, ideally USB 3.0 on the machine itself. Connect directly rather than through a hub or dock. Wait for background transfers, scans and updates to finish before copying or testing.
Check free space. Above 80 percent full, move data off and retest. Check the file system too. FAT32 adds overhead and caps single files at 4 GB. exFAT is usually the right choice for large files on removable drives across Windows and macOS. NTFS gives permissions and large file support on Windows at the cost of journaling overhead. (NTFS vs FAT32 vs exFAT)
If you move large files regularly and the drive cannot keep up, that is a hardware ceiling. Look for a drive with an explicitly stated sustained write speed rather than a headline read number. Reputable vendors list both.
Should you format a slow USB flash drive? Formatting is not a speed fix. Formatting clears file system clutter and can help if the file system is damaged or badly fragmented, but it does not change the flash memory, the controller or the port. A slow drive formats into a slow drive.
1. How can I test USB flash drive speed for free?
Every method in this article is free. For a rough number, copy a large file and time it. For comparable results, use a benchmark tool. DiskGenius includes one in the free version.
2. Why is my USB 3.0 flash drive so slow?
Check the port first. A USB 3.0 drive in a USB 2.0 port runs at USB 2.0 speed, and hubs or adapters do the same thing. If the port is fine, look at the drive's own write specification, how full it is, and whether you are copying many small files. Also keep in mind that USB 3.0 describes the connector rather than the flash inside, so a budget drive can be legitimately slow.
3. Does benchmarking shorten a flash drive's lifespan?
Flash memory has a finite number of write cycles, so any writing uses some of that budget. A speed test uses very little compared with normal use. Testing a drive a handful of times is nothing to worry about.
4. How can I tell if my USB port is 2.0 or 3.0?
Blue plastic inside the port is a common sign but not reliable. In Device Manager, expand "Universal Serial Bus controllers" and look for "USB 3.0" or "xHCI" entries, as opposed to "Enhanced Host Controller" for USB 2.0.
DiskGenius is a one-stop solution to recover lost data, manage partitions, and back up data in Windows.
Download
Facebook
YouTube
X