What Is a Good Reaction Time?

The first time I saw my own number on a reaction test, I felt cheated. 248 milliseconds. A quarter of a second to do nothing but click a green box? Surely I was faster than that. I wasn’t. Almost nobody is. And once I understood why, that number stopped feeling like a failure and started feeling kind of incredible, because a lot has to happen inside it.

So let me answer the question straight, then explain the asterisks.

the honest bands

On a standard visual click test, here’s how I’d grade what you see:

Your time	What it means
Under 150 ms	Almost certainly anticipation, not reaction. You jumped.
150–200 ms	Genuinely fast. Pro-gamer, F1-grid territory. Rare.
200–250 ms	Quick. Better than most people who test.
250–300 ms	Normal. The big middle of the bell curve.
300–400 ms	A bit slow, or you were tired / distracted / on a laggy device.
400 ms+	Usually something’s off: old monitor, trackpad, half-asleep.

The median online reaction time is 273 ms, with a mean of 284 ms. That’s from over 81 million clicks on Human Benchmark’s public stats, a large self-reported online aggregate, not a controlled study, but about as solid a “typical” as we have. If you land near there, you are dead average. That’s not an insult. It’s most of humanity.

Now here’s the thing I keep having to correct people on. I’ve watched competitors and forum posts call 500 ms a “good” reaction time. It isn’t. 500 ms is half a second. On a simple visual test that’s slow, the kind of number you get when you’re exhausted, clicking with a trackpad, or running an old display. A “good” simple reaction is under 250 ms, and a genuinely fast one is under 200 ms. Don’t let anyone move that goalpost.

why your screen lies to you a little

Lab tests of simple visual reaction usually land around 200–250 ms. Web tests run a touch slower: they add roughly 10 to 50 ms of display and input lag on top of your actual biology. Your monitor has to draw the frame, your mouse or trackpad has to report the click, the browser has to register it. None of that is your nervous system. So if your web score is a bit higher than the textbook lab figures, some of that gap is hardware, not you. I cover the device-lag math in more depth on our average reaction time hub if you want the receipts.

The opposite is also true. If you’re scoring sub-150 ms on a click test, you’re not superhuman. You’re predicting the timer. About 100 ms is the practical floor for a simple human reaction. World Athletics treats any sprint start under 0.100 s as a false start by rule, because below that you physically could not have reacted to the gun. The sprint-start case is the one real exception: in the lab, Pain & Hibbs (2007) measured elite auditory sprint-start reactions below 100 ms, with the neuromuscular component under 85 ms. And even that’s the absolute ragged edge. There’s no cleanly verified single “record” lower than that. So that viral “0.04 second Bottas” F1 figure? Anecdotal. That’s anticipation off the lights, not a measured reaction. Treat it as a meme, not a benchmark.

it depends: on you, and what you’re testing

“Good” isn’t one number, because three things move it.

Age. Reaction time peaks in your early 20s (around 258 ms on web tests), then slowly drifts up. The MindCrowd study (n=75,666) found roughly 7 ms of slowing per year. The curve decelerates rather than adding up in a straight line, but it puts people in their 70s roughly 40–60% slower than the early-20s peak. Kids are slower than young adults too; the curve is U-shaped. So a 280 ms at 45 is a genuinely better result than a 280 ms at 22.

Modality. What sense you’re reacting to matters a lot. Sound beats sight. Auditory reactions run roughly 25–50 ms faster than visual ones (in one study about 284 ms for audio versus 331 ms for visual) because sound transduces faster than light does through the retina. Tactile is faster still. So comparing your audio reaction time score to your visual one isn’t apples to apples.

Complexity. The moment you add choices, everything stretches. Hick’s law: roughly 200 ms base plus about 150 ms per bit of decision, though those numbers are illustrative and vary a lot across studies (Proctor & Schneider 2018), not settled constants. Simple reaction lives around 150–300 ms. Choice reaction jumps to 300–700 ms. That’s not you getting worse. It’s the decision step doing real work.

where the time actually goes

Inside that quarter second: light hits your retina and relays toward visual cortex in about 50–100 ms. Then the big chunk: the central decision. Then the motor command fires down the nerve and the muscle moves. That middle decision step is the largest piece, and the most trainable one. It’s why gamers improve. Esports pros sit around 170–190 ms, regular gamers near 200 ms, casual players around 230 ms. The hardware in their heads isn’t special: the warm-up and the practice trimmed the decision.

And the boring stuff matters. Sleep deprivation slows you and makes you inconsistent. Caffeine, around 5 mg/kg, can partly claw that back. The effect is modest and varies between people, and it works mostly by sharpening attention (McLellan et al. 2014). So if your number looks bad, ask whether you’re tired before you blame your genes.

If you’re chasing a target, here’s mine: under 250 is good, under 200 is fast, and consistency beats one lucky click. Go run a few clean rounds on the main test, and if you want to know what the bar actually is for FPS, check reaction time for gaming.

Sources

Written by Lokesh Rathore, Founder of ReactScore. Lokesh Rathore is the founder of ReactScore and the person behind everything on it. Reflexes and reaction time are a genuine fascination of his, so he built the test he wanted to use himself: one that measures honestly, corrects for the lag your screen and mouse add, and explains the numbers instead of hiding them. He keeps every figure on the site grounded in public datasets and published research. More about the author · how we measure →

the honest bands

why your screen lies to you a little

it depends: on you, and what you’re testing

where the time actually goes

Sources

Test it yourself