Mouse DPI Test — Measure the Counts Per Inch
Turn on JavaScript to run the measurement — the page counts pointer movement as it happens.
This one needs a mouse. A touchscreen reports finger positions rather than mouse steps, so there is no count to divide. Open this page on a computer with a mouse plugged in.
1. How far will you move the mouse?
Measure it on your desk with a ruler and mark the start and end. The measurement is only as good as this number.
2. Drag exactly that far
Click here, then move the mouse in a straight line for the distance above. Click again or press Esc to finish.
0 counted so far
3. Your result
No run yet.
— DPI
- Counts reported
- —
- Display scale applied
- —
Runs
Average —
A browser cannot read the DPI setting out of your mouse. There is no such value to ask for, so every page that shows one either measures it the way this one does or looks up your model in a table. It also cannot tell you which of the mouse's stored steps you are on, or what the sensor is capable of — only how far the pointer moved while your hand moved a distance you measured.
About the Mouse DPI Test
DPI is the one mouse number that the mouse will not tell you. Nothing in the browser can ask a mouse what it is set to, and the operating system does not know either — it receives steps and moves the pointer, without ever being told how far those steps are supposed to be.
It has to be measured. Move the mouse a distance you have actually measured on the desk, count the steps that arrive, and divide. That is the whole method, and it is what this page does: you supply the ruler, the page supplies the counting.
The part everyone gets wrong is pointer speed. If the system is scaling movement before the browser sees it, you are measuring your settings rather than your mouse. This page asks the browser for unaccelerated movement whenever it can, and tells you which of the two it got — because the same number means different things in each case.
One drag is never straight, so do three and use the average. Two runs that disagree by a few percent are normal; two that disagree by half are telling you something moved that should not have.
How to get a number you can trust
- Put a ruler on the desk and mark a start and end point for the distance you picked above. Sticky notes work. Guessing the distance is the fastest way to a wrong answer.
- Click the pad, then move the mouse from mark to mark in one straight, unhurried motion. Speed does not matter to the count, but lifting the mouse does — keep it down the whole way.
- Click again or press Esc at the end mark. The number appears only then, because a running figure would swing wildly on the first few steps.
- Read the mode badge next to the result. If it says raw movement, your pointer speed settings did not touch the number. If it says OS-adjusted, fix those settings first and measure again.
- Do it three times and use the average. Compare that against the DPI your mouse is set to: within a few percent is a match, and a result near a round number like 800 or 1600 usually means you have found the step you are on.
Frequently asked questions
Is the Mouse DPI Test free, and does anything get installed?
It is free, needs no account, and installs nothing. The measurement happens in the page as you move: the browser reports pointer steps, the page adds them up and divides by the distance you typed. Nothing about your mouse or your movement leaves the browser, and closing the tab discards the runs.
How is this different from a polling rate test?
They measure two different properties and neither implies the other. Polling rate is how often the mouse reports, in reports per second, and it is about timing. DPI is how much movement each report describes, in counts per inch, and it is about distance. A 1000 Hz mouse can be 400 DPI or 3200 DPI, and this page cannot tell you the polling rate any more than that one can tell you the DPI.
Why is my result not a round number like 800?
Because your hand is not a machine. A drag that wanders slightly, starts a millimeter early or ends late lands a few percent off, and that is normal. Take three runs and look at the average: if it sits within a few percent of a common value, that value is almost certainly your setting. A result that is off by half, or that changes wildly between runs, is a sign that pointer acceleration is still on.
Do I need to change any system settings first?
Only if the badge says OS-adjusted. In that mode the system scales the movement before the browser sees it, so pointer speed must be at the middle notch and Enhance pointer precision must be off, or you are measuring the settings instead of the mouse. When the badge says raw movement the browser bypassed all of that and you can leave your settings alone.
Can it tell me my mouse's maximum DPI or its sensor?
No, and no page can. This measures the step you are currently on, not what the hardware is capable of and not what model it is. A page that names your sensor is matching your browser's self-description against a table of known devices, which is a guess that works on popular models and fails quietly on everything else.
Which distance should I pick?
Longer is better, so use 20 cm if your desk and mouse pad allow it. The error comes almost entirely from where you start and stop, so a fixed millimeter of slop matters half as much over twice the distance. Ten centimeters is a reasonable default and four inches exists for anyone whose ruler is marked that way.