Reaction Time Test: How Fast Do You Actually Respond?
Wait for the panel to turn green, then click or tap as fast as you can. Five attempts, then your average in milliseconds — plus where that sits against everyone else.
Your attempts
What this reaction time test measures
This is a simple visual reaction task: one signal, one response, nothing to decide. The gap it records is the time your nervous system needs to notice a change in light, route that through the visual cortex and motor system, and get a finger moving. That chain runs roughly a quarter of a second, and almost none of it is under conscious control.
Five attempts are used rather than one, because a single go is mostly luck. The test will not let you cheat it either — clicking before the green appears voids that attempt instead of recording an impossibly fast time. If you would rather stop sooner, one attempt is enough to unlock the finish button.
However you found it — as an online reaction time test, a reaction tester, a reaction speed test, or a reaction time calculator — it does the same job, and it needs nothing but a browser. Use it to check reaction time before and after a coffee, or to test my reaction time against a friend on the same laptop; a like-for-like reaction time check on identical hardware is by far the most meaningful comparison you can run.
Average human reaction time
The average human reaction time for a simple visual cue lands somewhere between 250 and 300 milliseconds, and 273 ms is the number that gets quoted most often from large online samples. Expressed as an average reaction time in ms it is a single figure; expressed honestly it is a distribution with a long tail to the slow side. Very few people are dramatically faster than the middle, but plenty are much slower on any given attempt.
Search terms vary more than the underlying fact — average reaction time ms, normal reaction time, average human reaction time in ms all point at the same ballpark. The phrase human benchmark reaction time gets used loosely for whatever median a large online sample reports, and the ~273 ms figure this page compares you against comes from that same body of aggregated browser tests.
| Result | Reading |
|---|---|
| Under 200 ms | Exceptional for a browser test — check for an early click. |
| 200–250 ms | Fast. Common among gamers and younger testers. |
| 250–300 ms | The middle of the distribution. |
| 300–350 ms | Slower than average, often tiredness or hardware. |
| Over 350 ms | Worth retesting when rested, on a wired mouse. |
Laboratory measurements come out faster than any of these, frequently in the 190–250 ms band, because they cut out the display and input delays that browser testing cannot escape.
How the percentile is estimated
Human reaction times are not symmetrically distributed. There is a hard floor — nerve conduction and muscle activation cannot be rushed below roughly 150 ms — but no ceiling, so the curve leans heavily to the slow side. Modelling it as a log-normal distribution centred on 273 ms captures that shape well enough for a rough comparison, and that is what produces the "faster than ~X%" figure above.
Two caveats that matter. It is an approximation, not a measured rank against a real database of results. And it is calculated on the raw number your browser recorded, which already contains 10–50 ms of hardware latency — a wireless mouse on a 60 Hz screen can cost you a full percentile band compared with a wired mouse on a 144 Hz one. Treat the percentile as a broad bracket rather than a score.
Reaction age
Reaction speed declines gently and predictably with age, which makes it easy to run the comparison backwards: given your average, which age band does it most resemble? Our scale sits near 218 ms in the early twenties, drifts up slowly through the thirties and forties, and reaches roughly 310 ms by 70.
Those anchors describe the reaction alone, whereas your score also contains whatever your screen and mouse added, so 40 ms comes off before the comparison is made. Skip that step and the scale contradicts itself — a dead-average result would land on the 50th percentile and a reaction age near sixty at the same time.
The decline is real but gradual — around ten milliseconds per decade up to 40, steepening sharply after that — and it is dwarfed by day-to-day factors. Sleep loss, alcohol, illness and simply being cold will each move your result further than twenty years of ageing. So a reaction age well above your real one is a comment on today, not a diagnosis; run it again after a proper night's sleep before drawing any conclusions.
Other ways people measure reaction time
The green-box format is the most common, but it is not the only one, and the variants measure meaningfully different things:
- Driving reaction time test — a brake light replaces the green panel and the result is converted into stopping distance. A driving reaction time test measures the same simple visual response; the realism is in the presentation. Real emergency braking also involves recognising a hazard, which takes considerably longer.
- Auditory reaction time test — respond to a beep instead of a colour change. An auditory reaction time test typically returns times 20–40 ms faster than a visual one, because sound reaches the brain through a shorter processing path.
- Ruler drop test — the classroom classic, and the only one needing no screen. In the reaction time ruler drop test a partner releases a ruler between your open fingers and the distance it falls before you catch it converts to time. It costs nothing and has no software latency, though catching is a different motor task from clicking.
- Themed browser tests — parking and driving-branded versions such as the just park reaction time test are popular in the UK and work identically underneath, cue and click.
Whichever format you use, the rule for comparison is the same: only compare results measured the same way, on the same hardware, in a similar state.
What moves your number
Sleep is the biggest lever by a distance — a week of six-hour nights degrades response speed measurably, and unlike sleepiness itself, that decline does not feel like anything. Caffeine helps for a few hours. Being physically warmed up helps. Alcohol hurts sharply, and so does the tail end of a long working day.
Equipment is the second big one. Wired mice beat wireless, high refresh rates beat 60 Hz, and touchscreens are slower than either. Practice produces genuine improvement on this exact task within a few dozen attempts, mostly by teaching you not to over-anticipate — but do not expect it to carry over to anything else.
Frequently Asked Questions
-
For a simple visual cue like this one, large online samples cluster around 270–280 ms, and 273 ms is the figure most often quoted. Laboratory setups measure faster — often 190–250 ms — because they remove the screen and input delays a browser test cannot avoid.
-
Under 250 ms is quick for a browser test, 250–300 ms is ordinary, and over 350 ms usually says more about tiredness, a laggy screen or a trackpad than about you. Trained athletes and habitual gamers cluster near the fast end, but the spread within any group is wider than the gap between groups.
-
Attention wanders in the seconds before the signal, and that variation is a real part of reaction speed rather than noise to be discarded. This is exactly why the test runs five attempts and reports the average — a single lucky click tells you very little.
-
Your true reaction is faster than the number shown. Between the display refreshing, the panel actually changing colour, and the mouse or touchscreen reporting your click, somewhere around 10–50 ms of hardware and software delay is baked into every result. It is consistent enough for comparing yourself against yourself, and rough for anything else.
-
Somewhat, and the cheap wins come first: sleep, caffeine timing, and simply being warmed up rather than cold. Practice on a specific task improves that task quickly, though it transfers poorly to unrelated ones. The slow decline that comes with age is real but modest compared with how much a bad night's sleep costs you.
-
A themed variant that swaps the green box for a brake light and adds stopping distance to the result. The measured quantity is the same simple visual reaction; the extra realism is in the framing rather than in the measurement.