Average Reaction Time: How Fast Is Normal? (2026)

July 19, 2026 · By ClickBattle

Ask what the average reaction time is and you'll get four different numbers from four different pages, none of which mention that they're measuring different things. Some quote 250ms. Some say 200. Some cite an online test's median and call it the human average. They're not all wrong, exactly. They're just answering slightly different questions and not telling you which one.

The average human reaction time to a visual signal is roughly 200 to 250 milliseconds — about a quarter of a second. Laboratory studies using dedicated equipment cluster nearer 190ms, while large online datasets read closer to 273ms. The gap is measurement, not biology.

So this is the version with the asterisks left in. Where the numbers actually come from, why sound beats sight, how the figure drifts across your lifespan, and the distinction between the two kinds of reaction time that most pages skip entirely. If you're mainly here to find out whether your number is any good, that's a different question and we answer it properly in what counts as a good reaction time — this page is about the average itself.

What is the average reaction time?

There isn't one number, and any page that gives you one without a caveat is rounding off something important.

The most-cited laboratory figure for simple visual reaction — a light appears, you press a button — sits around 180 to 200ms. Robert Kosinski's literature review at Clemson University, which pulls together decades of reaction-time studies, puts it at about 190ms for a light stimulus. That's the number from controlled conditions: purpose-built equipment, a dedicated response button, and no browser anywhere in the chain.

Then there's the web figure. Human Benchmark, which has collected one of the largest public reaction-time datasets in existence, reports a median around 273ms. That's roughly 80ms slower than the lab number, and it's the figure most people actually encounter, because most people meet this topic through an online test rather than a research paper.

Both are real. Neither is wrong. The lab number tells you what a human nervous system can do when nothing is in the way; the web number tells you what a human plus a monitor plus a mouse plus a browser does. Since you probably play games on exactly that stack, the web number is arguably the more useful one for you — it's just not the number your biology is capable of.

Which is why "about a quarter of a second" is the honest headline. It's the range that survives both methods, and it's close enough for every practical purpose. If someone tells you the average is precisely 250ms, ask them measured how.

One thing worth setting down early: reaction time is its own scale and it doesn't map onto anything else. It's not clicking speed, and a fast clicker isn't automatically a fast reactor — those are separate skills measured in different units, which is why a good CPS and a good reaction time are two unrelated conversations. Our rundown of every speed test lays out which test measures which thing if you're trying to work out where to start.

Why is auditory reaction time faster than visual?

Because sound gets to your brain quicker. That's genuinely most of it.

The figures are consistent across the literature. Kosinski's review puts simple auditory reaction at 140 to 160ms and tactile — reacting to touch — at around 155ms, both comfortably ahead of the 180 to 200ms for vision. Sound wins, touch is close behind, and sight comes last.

That ordering surprises people, because vision feels like the sense you'd trust in a hurry. But the delay isn't in the reacting; it's in the arriving. An auditory stimulus takes roughly 8 to 10ms to reach the brain. A visual one takes 20 to 40ms. Before you've had any chance to respond, sight has already spent an extra 20-odd milliseconds getting there.

The reason comes down to how each sense converts the world into a signal. Your ear does it mechanically — sound moves the eardrum, the movement travels through bone, and hair cells in the cochlea turn that motion into nerve impulses almost immediately. Your eye does it chemically. Light hits the retina and sets off a cascade of molecular reactions before a signal ever leaves for the brain. Chemistry is slower than physics, and that difference shows up in your reaction time.

There's a caveat worth keeping, though: the gap narrows if the visual stimulus is bright enough. Turn the intensity up far enough and the difference between reacting to light and reacting to sound can largely disappear. So it's not that vision is inherently a slow sense — it's that at ordinary intensities, it's carrying an extra transduction step.

Practically, this is why a starting pistol is a sound and not a flag, and why rhythm games feel more precise than they have any right to. It also means every visual reaction test you've ever taken — including ours — is measuring your slowest modality on purpose. That's the right choice, because screens are what you actually react to.

How does reaction time change with age?

It follows a long, gentle arc, and the decline is slower than people fear.

Reaction time sharpens through childhood and keeps improving into your late 20s, which is where it bottoms out. From there it increases slowly through your 50s and 60s — slowly enough that you'd struggle to notice it year to year — and then lengthens more steeply from the 70s onward.

Two things are worth pulling out of that curve. The first is how flat the middle is. Between your late 20s and your 50s, the change is real but small, which means a well-rested 45-year-old will comfortably beat a tired 22-year-old on any given morning. Age sets a gentle ceiling; your state on the day moves you around under it far more.

The second is that the age effect isn't uniform across tasks. It's more marked for complex reaction tasks than simple ones. Pressing a button when a light appears holds up well with age. Deciding which of several buttons to press, under time pressure, degrades faster. The extra step — the deciding — is the part that ages, more than the raw signal-to-muscle path.

Which brings us to the distinction that explains a lot of confusion in this topic.

What's the difference between simple and choice reaction time?

Most pages quote one number and never mention there are several kinds. The difference matters, because they're not close.

Simple reaction time is one stimulus, one response. Light goes green, you click. Nothing to decide. This is what almost every online test measures, and it's where the 180 to 250ms figures come from.

Recognition reaction time adds a filter. Several things can appear and you respond only to certain ones, ignoring the rest. Now you have to identify before you act, and the cost is steep — one study Kosinski cites found simple reaction averaging 220ms while recognition reaction in the same setup averaged 384ms. Adding a single "is this the right thing?" step cost over 150ms.

Choice reaction time is the slowest of the three. Multiple stimuli, and a different response for each. You're not just detecting and filtering, you're selecting. More options mean more time, reliably.

This is the honest reason your test score doesn't predict how you'll perform in a game. When you click a green box, you're doing simple reaction. When someone peeks a corner and you have to decide whether to shoot, back off, or reposition, you're doing choice reaction — and that runs meaningfully slower for everyone, no matter what their green-box number says.

It also explains why experienced players seem to react faster than their measured reflexes allow. They're not beating biology. They've turned choices into non-choices through familiarity, collapsing a choice reaction back toward a simple one. That's a trainable thing, and it's most of what "game sense" actually is. The mechanical half — getting your crosshair onto the target once you've decided — is a separate trainable skill, and we cover that side in aim training for PvP.

Why do online reaction tests read slower than lab measurements?

This is the part worth reading carefully, because the usual explanation is wrong in an interesting way.

The ~80ms gap between the lab's 190ms and the web's 273ms isn't people being slower than science says. It's everything sitting between the signal and your finger that a lab bench doesn't have. And it's worth being precise about where that time actually goes, because one popular culprit is largely innocent.

The display is the biggest single contributor. Your monitor doesn't update continuously — it redraws on a fixed cycle. At 60Hz that's one refresh every 16.7ms, so on average you're waiting about half a frame just for the green to appear, and up to a full frame in the worst case. At 144Hz that interval drops to roughly 6.9ms, and at 240Hz to about 4.2ms. On top of the refresh cycle sits the panel's own processing and response time. A gaming monitor handles this in a few milliseconds. A television with image processing enabled can add tens of milliseconds on its own, which is why testing on a TV produces numbers that look almost comically bad.

Your input path adds a little more. A mouse reports its state at a fixed polling rate: 125Hz means an update every 8ms, 1000Hz means every 1ms. Then the operating system, the browser's event loop, and the compositor each take their small cut. Individually these are single-digit milliseconds. Stacked, they're not nothing.

The browser clock, though, is mostly exonerated — and this is the bit that gets misreported. You'll see claims that JavaScript timing is too coarse to measure reaction time properly. It isn't. Browsers did deliberately reduce timer precision after the Spectre vulnerability: Chrome clamps its high-resolution timer to 100 microseconds with a little random jitter added, where it used to offer 5. But 100 microseconds is 0.1ms. Against a reaction time of 250ms, that's a rounding error roughly two thousand times smaller than the thing you're measuring. The timer isn't your problem. The screen is.

So the honest summary: a browser test measures you plus your hardware, and the hardware tax typically runs somewhere in the tens of milliseconds. That's not a flaw to be corrected — it's the same tax you pay in every game you play on that machine, which makes the number genuinely representative of you in the situation you care about. It just isn't your nervous system's raw figure, and nobody should quote it as one.

Two practical consequences. Comparing your score to a friend's only means something if you're on broadly similar hardware. And chasing a faster number by upgrading your monitor is real — you'll genuinely see the improvement — but you haven't gotten faster, you've just stopped waiting as long. For how to read your own result with all this in mind, our reaction time test guide walks through what the number means once you have one.

What actually changes your reaction time?

Less than you'd hope, and the biggest lever is the least interesting one.

Sleep is the heavyweight. Nothing else on this list moves your number as much as being tired does, and it's the one factor that can shift you a whole band on a bad day. If you test first thing after a rough night and again after a proper sleep, you'll likely see the difference without needing a stopwatch to confirm it.

Warming up is free and undervalued. Your first attempt of any session is almost always your worst — cold hands, cold attention, and no feel for the timing yet. A handful of throwaway rounds before you take a score seriously removes a chunk of noise that has nothing to do with your actual reflexes.

Practice helps, at the margins. Regular testers do get faster, partly through genuine sharpening and partly by learning the task — specifically, learning to react cleanly rather than tensing up and jumping the gun. That second part is bigger than people assume. Anticipating the signal doesn't make you faster; it makes you post false starts.

Caffeine does something small. It's a real effect and a modest one. Treat the exact millisecond figures you'll see quoted around the internet as directional rather than precise — they come from varied conditions and doses, and they're not large enough to turn an average reactor into a fast one.

Lower-latency hardware works, but it isn't you. A faster monitor and a wired mouse will genuinely lower your measured score, for the reasons in the last section. Worth doing if you care about the number; just be clear you've reduced the tax, not raised the ceiling.

And the honest limit on all of it: your raw simple reaction time is largely set by biology, and you are not going to train a 280 into a 150. What you can do is stop giving away the milliseconds you're currently donating to fatigue, cold hands, and a slow screen. For most people that's a meaningful amount — and it's the only part actually under your control.

If you want your own figure to put against all this, the reaction time test takes about thirty seconds and averages several attempts, which matters more than it sounds: a single go can be a fluke in either direction, and the average is the only version worth quoting.