Reaction Time World Records

The honest answer first

There is no world record for reaction time. Nobody ratifies one, because there is no agreed apparatus, no agreed protocol, and no way to supervise an attempt made on somebody's own laptop. Any page presenting a specific name and number as the world record is presenting folklore.

What does exist is properly measured reaction data collected for other purposes, mostly in sport, under conditions strict enough to be worth something. That data is more interesting than a made up record anyway, because it shows where the real ceiling sits.

The 100 millisecond floor

Start with the limit, since everything else sits above it. A simple reaction requires light or sound to be transduced into a nerve signal, that signal to reach the relevant cortex, a motor command to travel back out, and a muscle to actually contract and produce force. Every one of those steps costs time, and synapses in particular are slow relative to the nerve fibre between them.

Added up, the floor for a simple response lands somewhere around 100 to 120 milliseconds. This is not a training limit. It is a transmission limit, and it is the reason the number appears in the rulebook of a sport rather than only in a laboratory.

Sprint starts, the best data anyone has

Athletics is the one place where reaction time is measured seriously, at scale, under supervision. Starting blocks in championship competition contain force sensors, and reaction is the interval between the gun and a defined rise in force against the blocks. The athlete cannot fake it, the equipment is standardised and the result is recorded for every competitor in every round.

World Athletics rules that a response quicker than 0.100 seconds is a false start, on the basis that a human could not have reacted that fast and must therefore have anticipated. The fastest legal starts in major championships cluster just above that threshold, in the 0.10 to 0.13 second range, with most finalists somewhere between 0.13 and 0.17.

The threshold is genuinely contested. Some research argues that a small number of people can beat 0.100 s legitimately, and that the rule therefore penalises the fastest reactors. The counterargument is practical rather than physiological: a single fixed line applied to everybody is defensible, and a line that moves per athlete is not.

Formula 1, and why the numbers look different

Published Formula 1 launch reactions generally fall between 0.2 and 0.4 seconds, with the very best around 0.15 s. Those look slow next to a sprinter until you notice they are measuring something else: the interval from the lights going out to the car actually moving. That includes the driver's reaction, the clutch release, and the physical response of a car with mass.

There is a second reason F1 numbers run slower. Detecting that something has switched off is marginally harder than detecting that something has switched on, and the five second light sequence loads the driver with anticipation first. You can feel both effects on our F1 start lights test, which reproduces the real procedure including the unpredictable hold.

Why online scores are not records

A browser cannot verify anything that matters. It does not know your monitor's refresh rate, your input latency, whether the tab was throttled, whether a script was modified, or whether a person was even involved. A page can only report the interval between two events it observed, and treat that as a measurement in good faith.

There is also the arithmetic of large numbers. On a test taken millions of times, some proportion of guesses will land within a few milliseconds of the signal purely by chance, and those land in the leaderboard rather than in the discard pile because the site cannot tell a lucky guess from a fast reaction. A public leaderboard on a test this easy to fake would be measuring persistence, not reflexes, which is why this site does not run one.

For what it is worth, on a visual click test a genuine 150 ms average is exceptional, 120 ms is at the edge of plausibility with excellent hardware, and anything under 100 ms is a false start rather than a result. Our reaction time test voids attempts made before the signal for exactly that reason.

If you want your fastest legitimate number

Use a wired mouse and the highest refresh rate screen you have, since that alone is worth 10 to 20 ms against a 60 Hz laptop. Be rested, warm and a few rounds into the session rather than cold on the first attempt. Sit close, keep your finger resting on the button rather than hovering, and look at the centre of the panel rather than scanning it.

Then take the average rather than the best of twenty, because the best of twenty is mostly a story about how many times you tried. If you want to see the auditory advantage, the sound reaction test is usually 20 to 40 ms quicker for the same person, and it is the closest thing here to how a sprint start actually works.