Mouse polling rate test showing 144, 165 or 500 Hz instead of 1000
If the number matches your monitor's refresh rate, the test is counting frames, not mouse reports. Chrome passes mouse moves to a web page once per screen refresh, so a simple tester reads 60, 144, 165 or 240. Only a tester that reads the batched reports inside each event, on an https page, can see 1000 Hz and up. Any other low number is a mouse-side problem, and the table below says which.
Your test is most likely counting browser events, not mouse reports. Chrome holds back mouse move events and hands them to the page once per screen refresh. So a simple web tester reads 60, 144, 165 or 240 Hz. That is your monitor's refresh rate, not the mouse. A tester can only see 1000 Hz and up if it reads the reports batched inside each event, and only on an https page. A reading near 500, or one that jumps around, is a different problem: slow movement, a low DPI, a setting, or the port.
What a polling rate test counts
Polling rate is how many times a second the mouse sends a report of its movement to the PC. At 1000 Hz that is one report every millisecond, as our polling rate guide works out. A polling rate test tries to count those reports.
A web page cannot see USB reports, though. It sees events: the messages the browser passes to the page's code when the pointer moves. A web tester counts events per second and calls that your polling rate. That is only right if the browser passes on one event for every report, and modern browsers do not.
Why the reading matches your monitor
Since Chrome 60, Google says, the input pipeline "will delay dispatching continuous events" such as mousemove and pointermove. It will "dispatch them right before the requestAnimationFrame() callback occurs". That callback runs once before each new frame is drawn. So Chrome sends at most one move event per refresh. Google calls this "frame aligned input", and did it for smoothness: in its tests the change cut hit tests by 35%.
The W3C Pointer Events standard allows the same in any browser. It says a move event may get "additional processing (for example to align with the display refresh rate)".
Nothing is lost, because the reports that arrived between two frames ride along inside the next event as coalesced events. Coalesced means merged: several reports packed into one event. A page gets that list by calling getCoalescedEvents(). A tester that counts the list sees the mouse. A tester that counts only the outer events sees the screen.
The two ways a web tester can see 1000 Hz
- getCoalescedEvents(). The list of reports inside each move event. It works in Chrome 58 and later, Firefox 59 and later, and Safari 18.2 and later, per MDN's compatibility data.
- pointerrawupdate. A separate event that skips the wait for the frame. The standard says browsers "SHOULD dispatch pointerrawupdate events as soon as possible and as frequently as the JavaScript can handle the events". It works in Chrome 77 and later and Firefox 148 and later. Safari does not have it.
Both need a secure context, which in practice means a page loaded over https. The standard marks both secure-only, and says pointerrawupdate fires "only ... within a secure context". MDN notes that Chrome before version 142 still sent pointerrawupdate on plain http pages. Current Chrome does not, so a tester at an http:// address falls back to plain move events and shows your refresh rate.
Even the raw event is not always one event per report. The standard says that if an earlier pointerrawupdate is still waiting, the browser "MAY coalesce" the new one into it. When the page is busy, reports get batched again. A tester that counts the list inside each raw event still gets them all.
Worked through: a 1000 Hz mouse on a 165 Hz screen
Take a mouse set to 1000 Hz, plugged into a rear port, with a 165 Hz monitor. Suppose you open a tester in Chrome that counts mousemove events, and move the mouse in fast circles. Chrome hands over one move event per frame, so the tester shows about 165. Each of those events carries about six reports (1000 / 165 = 6.06, derived). Now suppose you switch to a tester that reads getCoalescedEvents() on an https page. The same movement then adds up to about 1000 a second.
Next, suppose the mouse is set to 400 DPI and you move it slowly, 5 cm a second. That is 5 / 2.54 × 400 = 787 counts a second (derived). TechPowerUp explains that the set rate is only "the maximum applicable polling rate". If the mouse is not moved enough, "less updates will be transmitted". So this mouse cannot send 1000 reports a second, whatever the tester does. Move faster, or raise the DPI, and the count climbs back to 1000.
Read your number
| What the test shows | What it means | What to do |
|---|---|---|
| Your refresh rate: 60, 144, 165 or 240 | The tester counts frame-aligned move events | Use a tester that reads getCoalescedEvents() or pointerrawupdate, on an https page |
| About 125 to 133 | The mouse is on Bluetooth | Use the dongle or cable |
| Below the set rate, rising as you move faster | Not enough movement to fill every report | Move faster or raise DPI. 1000 Hz needs 1000 counts a second |
| A steady 500 however fast you move | The mouse is set to 500 | Check its software and any per-game profile |
| Jumps around, such as 500 to 1400 at 1000 Hz | Port, hub or a busy PC | Rear motherboard port, no other fast device on the same hub |
| Well under 8000 at 8000 Hz | Not enough movement, a USB 2.0 port, or the browser batching | Move faster at 1600 CPI or more, use a USB 3 port, confirm in a desktop tool |
The Bluetooth row comes from the link itself. Bluetooth LE before Core 6.2 cannot send more often than every 7.5 ms, which is 133 Hz, as our Bluetooth mouse polling rate page shows. A steady 500 points to the setting, because the set rate is a ceiling, so check the mouse's software and any per-game profile. For the jumpy reading, TechPowerUp advises a USB port native to the CPU, and no other high-polling device on the same hub. It adds that polling stays steady under load only if the system and OS are in good health. Our polling rate guide says the same about hubs and front-panel ports.
8000 Hz is the hardest to fill. At 1600 CPI, 8000 counts a second is 5 inches, or 12.7 cm, of movement every second (derived). TechPowerUp had to move the mouse "significantly more" to fill 8000 Hz than 4000. On a USB 2.0 (EHCI) port, "8000 Hz is never reached", which is why TechPowerUp recommends a USB 3 port in XHCI mode. Then the standard lets the browser batch raw events as well. So check an 8000 Hz reading in a desktop tool that reads raw input, not a web page. If a game stutters at that rate, that is a separate issue, covered in why an 8000 Hz mouse drops FPS.
Mouse fault or tester fault?
Two answers circulate. One says the mouse, the driver or the USB port is holding the rate back. The other says the test is wrong. Both are right, for different numbers: if the reading equals your refresh rate, it is the tester. A mouse fault would have to land on 144 or 165 by chance, and on exactly the rate of your screen. Any other low number is real and worth checking. The test is still useful, as long as it counts the right thing.
What this does not cover
Google documents frame alignment for Chrome. We found no matching statement for plain move events in Firefox or Safari. So read a Firefox or Safari result the same way: if it equals your refresh rate, the page counts outer events. A count of 1000 also says nothing about how evenly the reports are spaced, and spacing matters more than the average. One Firefox bug touched these events, but not the count: in Firefox 152 on Windows, pointerrawupdate in pointer lock added up too much movement at high polling rates. Mozilla marked it fixed in August 2026. And a browser number is not what a game sees. Games read the mouse their own way, not through a web page. Valorant, for one, reads mouse reports in batches, as our Raw Input Buffer page explains.
What the number means
- A reading equal to your refresh rate (60, 144, 165, 240) means the tester counts frames, not reports. Chrome sends move events once per frame.
- Only testers that read getCoalescedEvents() or pointerrawupdate, on an https page, can show 1000 Hz and up.
- Below the set rate but rising with speed: move faster or raise DPI. 1000 Hz needs 1000 counts a second.
- Steady 500: the mouse is set to 500. Jumpy: rear port, no shared hub. 125 to 133: Bluetooth.
- For 8000 Hz, move at least 12.7 cm a second at 1600 CPI, use a USB 3 port, and confirm in a desktop tool.
Sources
- W3C, Pointer Events Level 3, W3C Recommendation: section 4.2.5 "The pointerrawupdate event", section 10.1 "Coalesced events" and the PointerEvent IDL. Read via the web.archive.org snapshot of 18 November 2026.
- Dave Tapuska, Chrome for Developers, "Aligned input events".
- MDN, Element: pointerrawupdate event and browser compatibility data for getCoalescedEvents().
- Mozilla, Bug 2055058: pointerrawupdate in pointer lock on Windows at high polling rates.
- TechPowerUp, Razer Viper 8K review, "8000 Hz: A Closer Look" (26 March 2021).
Frequently asked questions
Why does my polling rate test show my monitor's refresh rate?
Because the test counts browser events, and Chrome sends mouse move events to a page once per frame. Since Chrome 60 it delays mousemove and pointermove and dispatches them right before the next frame is drawn. A tester that counts only those events reads 60, 144, 165 or 240, whatever your screen runs at.
Can a browser test show 1000 Hz or 8000 Hz at all?
Yes, if it reads the coalesced events inside each move event with getCoalescedEvents(), or listens for pointerrawupdate. Both only work on a secure (https) page. getCoalescedEvents() works in Chrome, Firefox and Safari 18.2 and later. pointerrawupdate works in Chrome 77 and later and Firefox 148 and later, not Safari.
Why does my 1000 Hz mouse only show about 500 Hz?
If the number rises when you move faster, you are not moving enough: the mouse only sends a report when it has movement to send, and 1000 reports a second need at least 1000 counts a second. If it sits at a steady 500 however fast you move, the mouse is set to 500 in its software or profile. If it jumps around, try a rear motherboard port with no other fast device on the same hub.
Why does my 8000 Hz mouse show 4000 or less?
8000 Hz takes a lot of movement to fill: at 1600 CPI you need about 12.7 cm a second. TechPowerUp also found 8000 Hz is never reached on a USB 2.0 (EHCI) port. And the standard lets a browser batch raw events when the page is busy, so confirm in a desktop tool before blaming the mouse.
Why does my Bluetooth mouse test at 125 or 133 Hz?
That is the Bluetooth link, not the tester. Bluetooth LE before Core 6.2 cannot send more often than every 7.5 ms, which is 133 Hz. Use the mouse's dongle or cable for a higher rate.