What Affects Your Reaction Time? Sleep, Caffeine, Age

The first time someone told me my reaction time was “just genetics,” I rolled my eyes. Partly true, partly nonsense. Yes, some of your reaction speed was handed to you at birth and you can’t return it. But a surprising amount of it is stuff you’re doing to yourself right now: last night’s sleep, the coffee you skipped, the cheap monitor you’re reading this on. So let’s actually separate the two. Here’s what affects your reaction time, which direction each factor pushes, and what you can do about it.

The starting point: on Human Benchmark’s public statistics (a large self-reported online aggregate) the median click test runs around 273 ms, with a mean of 284 ms across more than 81 million clicks. That’s a crowd-sourced figure, not a lab result. In a lab, without web lag, simple visual reaction lands closer to 200-250 ms. The factors below move you around inside that range.

age: you can’t change it, but it’s slower than you think

This is the big unchangeable one. Reaction time peaks in your early 20s (roughly 210-220 ms in a lab, around 258 ms on a web test) and then it drifts. The MindCrowd study (Talboom et al., 2021; n=75,666) put the slowdown at about 7 ms per year. Don’t just multiply that out. The curve decelerates rather than dropping in a straight line, but the net effect is that people in their 70s land roughly 40-60% slower than the 20s peak. Kids are slower than young adults too, so the whole curve is a U-shape: you climb up to your peak, plateau, and decline.

Seven milliseconds a year sounds tiny. Over decades it isn’t. But here’s the part people miss: the age curve is gentle enough that a well-rested, caffeinated 45-year-old can absolutely beat a sleep-deprived 22-year-old. Age sets your ceiling. It doesn’t decide where you actually land on a given day.

Same goes for the sex gap: in Der & Deary’s analysis (2006), men averaged about 30-34 ms faster than women on simple visual RT. That’s a group average, it’s smaller than the variation between individuals, and some studies report a narrower gap. Don’t read too much into it for any one person.

sleep: the biggest thing you control, and it cuts both ways

If you only fix one thing, fix this. Sleep deprivation slows your reaction time, and (this is the part that gets ignored) it makes you less consistent. Your mean creeps up and your worst clicks get much worse. Those random lapses, where you sit there for 450 ms like your brain went out for a coffee, are a sleep signature.

I’ve watched my own numbers swell after a bad night. It’s not subtle. So before you blame your genetics or your mouse, ask whether you slept. Then test yourself on a normal day and again after a rough night. The gap will probably embarrass you.

caffeine: a real fix, but a partial one

Caffeine helps, mainly by propping up attention. Around 5 mg/kg partly offsets the slowing from sleep loss (McLellan et al., 2014), though the effect is modest and varies from person to person. Note the word partly. It doesn’t restore you to a rested baseline. It mostly keeps your attention from collapsing, which is where sleep loss does its damage.

My honest take: caffeine is a patch, not a fix. Use it before a session if you want, but don’t pretend a double espresso replaces a night of sleep. It buys you attention, not new reflexes.

what modality you’re reacting to

Here’s a fun one you can’t change but should understand. Not all senses are equal. Tactile is fastest, then auditory, then visual. Sound beats light by roughly 25-50 ms. One study clocked about 284 ms for auditory versus 331 ms for visual. The reason is physical: sound transduces faster than light’s phototransduction in the retina, which eats up about 50-100 ms before your brain even gets the signal.

So if a test feels different from another, the modality might be why, not you. Compare yourself on the audio reaction test versus the visual one and you’ll likely be faster on audio. That’s expected.

how many choices you have

Simple reaction (one signal, one response) is the fast case, around 150-300 ms. Add choices and it balloons. This is Hick’s law: RT rises by a roughly fixed amount per bit of information. The usual textbook form is RT ≈ 200 ms + 150 ms per bit, but treat those numbers as illustrative. The actual constants vary a lot across studies. More options to sort through, more time. Choice reaction climbs to 300-700 ms.

That’s not a flaw. It’s the decision step, the largest and most trainable chunk of your reaction. Want proof of how much choices cost you? Run the choice reaction test right after a simple one and watch the number jump.

practice and fitness: the modest, real gains

Practice works, but be realistic about what it trains. It trims the decision step and your warm-up tax, not the hard physical floor. Light still takes 50-100 ms to reach your visual cortex no matter how good you are. Fitness helps modestly too. Neither turns you into a different human.

Factor	Direction	Can you change it?
Age	Slower with age (~7 ms/yr)	No
Sleep loss	Slower + less consistent	Yes
Caffeine	Faster (partial offset)	Yes
Modality (audio vs visual)	Audio ~25-50 ms faster	No (it’s physics)
Number of choices	Slower per Hick’s law	Sort of
Practice / fitness	Modestly faster	Yes
Device / display lag	Adds 10-50 ms	Yes

the factor nobody blames: your screen

Web tests add 10-50 ms of display and input lag on top of your true reaction. A laggy monitor, a wireless mouse with bad polling, a busy browser tab, all of it inflates your number. So a “bad” result might be your hardware, not you. Test on a wired mouse, close junk tabs, and you might quietly shave 20 ms.

One floor you cannot cheat: for a simple human reaction, about 100 ms is the practical limit. World Athletics flags any sprint start under 0.100 s as a false start. Elite sprinters reacting to a starting gun have been measured just below that in the lab. Pain & Hibbs (2007) found sprint-start auditory reactions under 100 ms, with the neuromuscular component under 85 ms, but that’s a trained athlete reacting to sound under ideal conditions. For an ordinary visual click test, if you ever come in under 100 ms, you anticipated. You didn’t react.

Stop guessing about your own number. Sort out sleep, fix your hardware, then read how to improve reaction time and retest. The honest comparison is the whole point.

Sources

• Human Benchmark: Reaction Time statistics (self-reported online aggregate: median 273 ms, mean 284 ms, 81M+ samples)
• Talboom et al. (2021), npj Aging: MindCrowd reaction-time study (n=75,666; ~7 ms/year age slowing)
• Der & Deary (2006), Psychology and Aging: age and sex differences in reaction time
• McLellan et al. (2014): caffeine (~5 mg/kg) and reaction time after sleep loss
• Proctor & Schneider (2018), QJEP: review of Hick’s law
• Pain & Hibbs (2007), Journal of Sports Sciences: sprint starts and the minimum auditory reaction time
• World Athletics: IAAF Sprint Start Research Project (100 ms false-start rule)

---