Click a reaction time test a few times in a row and you'll get a different number almost every attempt -- 230ms, then 190ms, then 260ms. It feels random, but it isn't. There's a real physiological process happening between the screen changing color and your finger moving, and that process has a floor that essentially no one can push below, no matter how focused or caffeinated they are.
What actually happens in those milliseconds
A simple visual reaction time test is measuring a chain of events, not a single action: light hits your retina, the signal travels up the optic nerve to the visual cortex, your brain recognizes the change and sends a motor command down to your hand, and your muscles finally fire to produce the click. Every link in that chain takes real time -- it's basic neural conduction, not a reflex you can shortcut with practice.
For most healthy adults, that whole chain resolves in roughly 200 to 250 milliseconds for a simple "click when you see it" test. People who train specifically on these tests can shave that down toward 150-180ms with practice, but getting meaningfully under 150ms on a genuine visual reaction -- as opposed to a lucky guess -- is vanishingly rare. The nervous system just doesn't move signals that fast.
The 100-millisecond rule that proves the floor is real
Competitive sprinting gives a clean, real-world illustration of this. World Athletics rules a sprinter's start a false start if they leave the blocks in under 100 milliseconds (0.1 seconds) after the gun fires -- not because it's against the rules to be fast, but because it's considered physiologically impossible to have actually reacted to the sound that quickly. Anything under that threshold is treated as anticipation, not reaction. Elite sprinters, who are about as primed and practiced as a human can be for a single reaction, still average closer to 120-160ms off the gun, and even they can't reliably beat 100ms through reaction alone.
That's useful context for reading your own score: if a browser-based test ever reports something like 80 or 90 milliseconds, you likely clicked a moment too early rather than reacting unusually fast.
Why your own scores bounce around so much
A few things move the number run to run, separate from the underlying floor:
- Attention lapses. If your mind wandered for even a fraction of a second before the stimulus appeared, you lose time catching back up, which shows up as a slower result.
- Fatigue and arousal. Being tired, understimulated, or even just bored slows reaction time measurably; being mildly alert (not over-caffeinated or anxious) tends to produce the fastest, most consistent results.
- Practice and anticipation. After a few rounds you start to anticipate roughly when the change will happen, which can shave real time off later attempts -- this is why a "personal best" after ten tries isn't a fair comparison to your first attempt.
- Hardware and software overhead. A browser test adds its own small, mostly fixed delay from display refresh rate (a 60Hz monitor redraws every ~16.7ms) and input polling, on top of your biological reaction time. This doesn't change dramatically between runs, but it means the number you see is "your reaction time plus some device overhead," not a lab-grade neurological measurement.
What a reasonable score actually looks like
As a rough frame: under 200ms is fast, 200-250ms is typical for an attentive adult, and 250-300ms+ usually means some combination of distraction, fatigue, or just an off attempt rather than a meaningfully different reaction speed. Running the test several times and looking at your median result tells you more than any single click does, since one outlier -- fast or slow -- is mostly noise.
If you want to see where you land, the Reaction Time Test runs entirely in your browser: it flashes a change on screen, times your click, and gives you an instant millisecond score with no sign-up or download required.