Reaction Time Test

Test your reaction time in milliseconds over five tries, with average, median and best, next to lab norms.

Tries

Click, tap or press Space to start

5 tries

Try 1
Try 2
Try 3
Try 4
Try 5

Measure how fast you react to a change on screen. The panel turns red, then green after a random wait; click, tap or press Space as soon as it does. After 5 tries you get your average, median and best time in milliseconds, and how much they varied.

The test runs in your browser. Finished sessions are kept on this device, separately for mouse, touch, pen and keyboard, so you can watch your times across sessions.

Reading your result

Average is the mean of your counted tries, and median the middle one (the mean of the two middle ones for an even count). One slow try pulls the average up but barely moves the median, so the median is the steadier figure to compare between sessions. Best is your fastest try, and spread the standard deviation: how far a typical try sits from your average. All are whole milliseconds.

In a 2015 lab study by Woods and colleagues of 1,469 adults aged 18–65, the mean simple reaction time on a computer test was 231 ms, and 217.8 ms for ages 18–24. Their own screen and mouse added a measured 17.8 ms; without it the mean was 213 ms. Human Benchmark, a popular web test, gives a median of 273 ms across its visitors and says a computer may add 10–50 ms. No source publishes a population distribution for web tests, so the tool gives no percentile.

How to use it

Click, tap or press Space or Enter on the panel to start. It turns red with “Wait for green…”, then green with “Tap!” after 1.5 s to 5 s, any time in between being equally likely, so the moment cannot be guessed. Each panel also changes its icon and its words, not only its colour.

Too soon: a press while the panel is still red. Too fast: a press within 100 ms of green, shown with its time. Neither is counted; the try starts again with your next press, and the summary says how many there were.
Missed: no response within 2 s of green is a lapse, and the try starts again.
Switching to another tab during a try cancels it, as the browser stops timing a hidden page accurately.
A key held down counts once, and a second finger on a touch screen is ignored. After a counted try, a press within 300 ms is ignored, so the second click of a double click does not start the next try.

Choose 3, 5 or 10 tries. A session in progress is not saved; a reload starts afresh.

How the time is measured

The clock starts at the first animation frame after green is put on the page: the frame that paints it. It stops at the timestamp the browser gives your click, tap or key press, which is taken when the input arrives rather than when the page gets round to handling it. Both come from the same high-resolution clock.

What the tool cannot remove is your hardware's own delay. A 60 Hz screen draws a new frame every 16.7 ms, so green can appear up to that much after it is due; the screen, mouse or touch screen then add their own latency. That is why web results run slower than lab ones and vary between devices, and why your history is kept apart for each input.

The 100 ms cut-off is the tool's rule: the eye, brain and hand cannot respond that quickly, so a press that fast was begun before green appeared. Studies set their own cut-off; Woods and colleagues counted responses from 110 to 1,000 ms.

Worked example

Five tries with one slow one

Tries (ms): 243, 219, 236, 402, 228
Result: Average 266 ms · median 236 ms · best 219 ms · spread ±77 ms

The 402 ms try lifts the average to 266 ms, while the median stays at 236 ms, close to the other four.

Frequently asked questions

The lab mean of 231 ms (218 ms for ages 18–24) is a fair reference for a mouse on a computer, less the delay your own hardware adds; a touch screen typically reads slower. Compare yourself with your own history on the same device more than with anyone else's number.

Early sessions can improve as you get used to the test and your device. Your history on this device shows whether your times are still falling or have settled.

A touch screen scans for fingers at its own rate and the phone processes the touch before the browser sees it, which adds delay a mouse does not have.

Seeing the change, deciding and moving takes longer than that, so a press that fast was a guess that happened to land just after green. Counting it would reward guessing.