Reaction time test: how many milliseconds are actually yours?
Reaction time test
Click to start
Wait for green, then click as fast as you can. Clicking early costs you the round.
Average reaction
—ms
About 100 ms of every result belongs to your mouse and your screen, not to you — which is why the same person scores better on a 144 Hz monitor.
- Best try—
- Worst try—
- Rounds done—
- Early clicks—
One gear check a month
MonthlyOne thing a month worth measuring before it fails properly — a mic that reset itself, a stick that drifts, a switch sending two clicks. Two sentences and the test.
Try also
What is a good reaction time in milliseconds?
The reaction time test measures the gap between the frame that lights the target and the moment your click reaches the page, which means a fixed slice of every result is equipment rather than you. Woods and colleagues timed 1,469 adults aged 18 to 65 and found a mean simple visual reaction time of 231 ms — 213 ms once the delay of their own apparatus was subtracted.
That 18 ms correction is the useful figure. A browser on consumer hardware carries more of it than a laboratory rig does, because the screen holds each frame for its full duration, the mouse reports on its own polling interval, and the operating system delivers the event when it next looks.
So read 200 to 250 ms across five rounds as ordinary, expect your single best round to sit well under your average, and treat anything below 150 ms as a click that started before the target did.
A reaction test cannot separate you from the screen in front of you, but frame time arithmetic says how much of the result belongs to each. A display shows one frame at a time, so a target that becomes visible has waited on average half a frame to be lit — before the mouse has done anything at all.
| Display | Frame time | Average wait to be lit | Worst case |
|---|---|---|---|
| 60 Hz | 16.7 ms | 8.3 ms | 16.7 ms |
| 75 Hz | 13.3 ms | 6.7 ms | 13.3 ms |
| 120 Hz | 8.3 ms | 4.2 ms | 8.3 ms |
| 144 Hz | 6.9 ms | 3.5 ms | 6.9 ms |
| 240 Hz | 4.2 ms | 2.1 ms | 4.2 ms |
Between a 60 Hz laptop and a 240 Hz monitor that is about 6 ms on average and 12 ms at worst, which is roughly a quarter of the 26 ms that the same research attributes to ageing from 18 to 65. Measure the panel you are actually sitting at with the refresh rate test, the mouse half of the chain with the polling rate test, and the pointing side of the same reflex with the aim trainer.
Frequently asked questions
Is 200 ms a good reaction time?
It is slightly better than average for an adult. The largest published sample, 1,469 people aged 18 to 65, put the mean simple visual reaction time at 231 ms, and that was on laboratory equipment with 18 ms of hardware delay already removed. On a browser, with a consumer screen and mouse in the chain, a genuine 200 ms is a good result. Below 150 ms is faster than the visual system normally manages and usually means an anticipated click.
Why is my score different on another reaction test site?
Three reasons, and all of them are worth 10 to 30 ms. Sites start the clock at different points, some when the colour change is requested and some when the frame carrying it is actually rendered. Stimuli differ: a colour change is spotted faster than a shape appearing. And browsers coarsen their timers against fingerprinting, with Firefox rounding event timestamps to 2 ms by default and to 100 ms if resistFingerprinting is switched on.
Does my monitor refresh rate change the number?
Yes, and predictably. A display shows one frame at a time, so a target that becomes visible has waited on average half a frame to be lit: 8.3 ms at 60 Hz, 2.1 ms at 240 Hz. Add the panel’s own response time on top. The same person tested on a 60 Hz laptop and a 240 Hz monitor will usually be 5 to 10 ms apart before the mouse has contributed anything.
Does the mouse matter, or is that marketing?
It matters, by a smaller amount than the screen. A 125 Hz mouse holds a press for up to 8 ms before reporting it, so it costs 4 ms on average; a 1000 Hz mouse averages 0.5 ms. Modern wireless receivers add roughly a millisecond and are no longer the weak link. Across the whole input path, expect 2 to 10 ms — real, but smaller than the difference between two people on the same equipment.
Why is my average so much worse than my best round?
Because attention lapses are not symmetrical: you can be 40 ms slower than usual but almost never 40 ms faster. Over five rounds a single distracted attempt at 500 ms adds 50 ms to the mean while the median does not move at all. That is why the best round is the number people quote and the average is the number that describes them.
Does reaction time really get worse with age?
It does, and more slowly than the internet suggests. The same 1,469 person study found simple reaction time rising 0.55 ms per year of age, which is about 26 ms across the whole 18 to 65 span. That is roughly four times the 6 ms you would gain by moving from a 60 Hz screen to a 240 Hz one — the ageing effect is real, but it is the same order of magnitude as the hardware you are testing on.
How the test works
The clock starts on the requestAnimationFrame callback for the frame that draws the target, and
stops at the event.timeStamp of your click. Both are high resolution timestamps measured from the
same page time origin, so the subtraction is exact even though the two numbers come from different
places. Rounds where the click arrives before the target appeared are discarded rather than scored
as zero, and the page reports the mean of the surviving rounds alongside the fastest one.
The limits are structural and no browser test escapes them. The frame we time from still has to be scanned out to the panel, which takes up to one full frame period; the panel then needs its own response time; the mouse reports the press on its polling interval; and the operating system delivers the event on its next input tick. Firefox additionally rounds event timestamps to 2 ms by default as a fingerprinting defence. Every one of those adds to the measurement, so a browser result is systematically higher than a laboratory one for the same person.
Worked example
With five reaction times (ms) 284, 241, 268, 255, 301, this page works out 268 ms. The median is what the tool reports, not the best attempt: one lucky round says nothing, and the middle value is the one that survives a bad round too.
- Median268 ms
- Best round241 ms
- Worst round301 ms
Sources:
Why the number moves between attempts
Simple reaction time is one number made of two jobs. Detecting that something changed takes most of the time; sending the click takes the rest, and only the second half improves with practice on a specific test.
Five rounds is the minimum that means anything. A single round is dominated by whether you happened to be looking at the right place; the spread across five tells you more than any one of them does.
Everything that affects attention affects the result. Time of day, sleep, caffeine and how many other tabs are doing work all move it by more than the difference between two monitors.