What polling rate is, and what it is not
Your mouse does not stream its position continuously. It reports on a fixed schedule, and polling rate is the frequency of that schedule. A 125 Hz mouse reports 125 times a second, once every 8 milliseconds. A 1000 Hz mouse reports every millisecond. Between reports, your computer simply does not know the mouse has moved.
That interval is the entire effect. A higher polling rate means your cursor position is more current and your click is registered sooner, by an amount bounded by the interval itself. Going from 125 Hz to 1000 Hz removes up to 7 milliseconds of input delay. That is real, and it is also the whole of what changes.
Polling rate is not DPI and the two are routinely confused. DPI is sensitivity: how far the cursor travels per inch of hand movement. Polling rate is timing: how often the mouse checks in. Raising DPI makes the cursor move further; raising polling rate makes it move sooner. They are independent, and neither compensates for the other.
It is also not the same as sensor tracking quality or click latency, both of which are separately specified and both of which can matter more than polling rate on a given mouse.
When you can actually feel the difference
The jump from 125 Hz to 500 Hz is the one most people notice, and mainly during fast movement, where a low report rate makes the cursor path look faintly stepped rather than smooth. From 500 Hz to 1000 Hz the improvement is smaller but still reported by a lot of people in fast-paced games.
Above 1000 Hz the returns get small quickly. The step from 1000 Hz to 8000 Hz removes at most 0.875 milliseconds of delay, which is comfortably below the frame time of any display you are likely to own, and well below the display and panel latency covered on the reflex tests. Some players report feeling it. The arithmetic says there is not much left to feel.
There is also a cost worth knowing about. Very high polling rates raise CPU load, because the system is handling eight times as many reports, and some systems handle that badly enough to lose frames. If you set 8000 Hz and your game gets less smooth rather than more, that is the likely reason. Test it rather than assuming higher is better.
How to check and change yours
To check it, run the polling rate test and move the mouse quickly and continuously inside the box for the whole run. The reading matters less than the method: read the next section before drawing a conclusion from a low number.
To change it, use the mouse manufacturer's own configuration software, where polling rate is usually a setting called report rate or polling rate. A mouse with no configuration software is almost certainly running at 125 Hz, which is the USB default, and there is usually no way to change it. Some mice also expose a hardware button or key combination that cycles the rate without software.
If you are picking a setting and do not want to think about it further, 1000 Hz is the right default for gaming and 125 Hz is genuinely fine for everything else. Neither choice is one you will regret.
Why a browser test always reads a bit low
This is the part that generates the most confused messages, so it is worth being precise. A web page can only measure the mouse events the browser hands it, and the browser delivers those in batches aligned to the display's frame rate rather than the instant each report arrives. A page therefore measures a floor: the true polling rate is at least the measured figure, and usually a little higher.
The measurement also only approaches the true rate while the mouse is moving continuously and fast. If you move it in short bursts, the gaps between movements are counted as gaps between reports, and the average collapses. A reading well under a standard value nearly always means the movement was not sustained, not that the hardware is underperforming.
So read the result this way: if the test shows a maximum near 1000 Hz, your mouse is a 1000 Hz mouse, even if the average sits lower. If the maximum never gets near it after several fast, continuous runs, then check the setting in the manufacturer's software before suspecting the mouse.
Frequently asked questions
What is a good mouse polling rate?
1000 Hz is the sensible default for gaming and the setting most people should pick. 125 Hz is genuinely fine for browsing and office work. Rates above 1000 Hz reduce delay by well under a millisecond and raise CPU load, so treat them as something to test rather than an automatic upgrade.
Is polling rate the same as DPI?
No. DPI is sensitivity, how far the cursor moves per inch of hand movement. Polling rate is timing, how often the mouse reports its position. They are independent settings and neither substitutes for the other.
Why is my measured polling rate lower than my mouse is rated for?
Because a browser measures a floor rather than a ceiling. It can only count the events it is handed, and it receives them in batches tied to the display's frame rate. Move the mouse fast and continuously for the whole run, and read the maximum rather than the average.
Does a higher polling rate use more CPU?
Yes. At 8000 Hz your system processes eight times as many reports per second as at 1000 Hz, and some systems handle that badly enough to lose frames. If raising the rate makes things feel worse rather than better, that is why.