Average reaction time, and how much of your score is the device
Updated 2026-10-04 · roughly a 4 minute read · take the test first so these numbers mean something.
Search "average reaction time" and you will get a single confident number, usually around 250 milliseconds. That number is real but incomplete in two ways that matter when you score yourself: it describes a laboratory setting, and it quietly ignores the screen you are holding.
What "reaction time" is being measured
A simple reaction time task has one stimulus and one response: see green, press. That is the 250 ms figure's home turf. Most real situations — a passing car, a ball, an opponent peeking a corner — are choice reaction time, where you first have to work out which response applies. Choice tasks routinely add 100–200 ms on top, because the delay is mostly decision, not nerve.
So a 270 ms score on a single-stimulus test and a "slow reaction" feeling in a game are not contradictions. Games reward predicting what happens next, which lets you start moving before the stimulus is fully there. Raw speed is the smaller half of the equation.
Practical bands for a simple visual test
| Average over 5 rounds | Band | Typical profile |
|---|---|---|
| Under 200 ms | Elite | Rare. Frequently reported by competitive FPS and fighting players. |
| 200 – 250 ms | Fast | Regular action-game players, younger adults, good sleep. |
| 250 – 300 ms | Average | Where most adults land on a first attempt. |
| 300 – 400 ms | Slow | Tired, distracted, cold hands, or a laggy touch screen. |
| Over 400 ms | Sluggish | Retest after sleep before drawing conclusions. |
These are working bands for a browser test, not clinical norms. If you have a genuine concern about reaction speed — medication effects, head injury, a sudden drop — that is a conversation for a clinician, not a web test.
The human reaction time average, by age
Reaction time rises through the decades, and the rise is steeper than most people expect once you get past your twenties. Roughly, in a simple visual task:
- Teens to mid-20s — the fastest window; scores in the low-to-mid 200s are common.
- Late 20s to 40s — a slow upward drift; the 250–300 ms band gets crowded.
- 50+ — the drift accelerates, and variance between people widens sharply. Two 65-year-olds can differ by more than a 20-year gap does.
The measured figures behind those bands come from one peer-reviewed study: Woods et al. (2015) tested 1,469 adults aged 18–65 on 120 trials each and recorded 217.9 ms for the 18–24 group rising to 239.1 ms for the 59–65 group — a whole-sample average of 231.0 ms and an age trend of about 0.55 ms per year, roughly 21 ms across the whole of adult life. That is less than a slow screen costs you. The reaction time test page prints all seven age groups with their sample sizes; the table is not repeated here on purpose.
Twenty-one milliseconds separates the youngest group from the oldest. The study's age trend is about 0.55 ms per year, so the whole of adult ageing moves the number less than the 10–40 ms your display and input path do. If your screen is slow, that is the bigger problem, not your birthday.
People differ more than age groups do. The participants' averages spread about ±27 ms around the study mean — larger than the 21 ms gap between age groups. A score 40 ms off the average for your band is ordinary, not a finding.
Ageing slows the finger, not the eye. In the same data the sensory-decision component held steady at about 131 ms across every age group, while the motor part got slower. Detection stays sharp into your sixties; the last centimetre of the movement is what costs you. (The study also found no sex difference in simple reaction time.)
The study stops at 65 and this page stops with it. Reaction time keeps drifting upward after that age faster than a straight line predicts, so any figure printed for a 75-year-old would be invented rather than measured; under-18s get the same honesty, since the study measured adults.
The device tax: 10 to 80 milliseconds you cannot train away
Every browser reaction test measures "paint to tap", and both ends of that interval are hardware:
| Stage | Typical added delay | Notes |
|---|---|---|
| Display refresh | ~4 ms (144 Hz) to ~17 ms (60 Hz) | Your tap can only be seen at the next refresh boundary. |
| Touch sampling & OS input | ~10 – 40 ms | Phones vary a lot; older or budget panels are worse. |
| Browser compositing | ~1 frame | Small here — this site changes a background colour, nothing heavy. |
| Wireless controller / BT keyboard | +10 – 30 ms | If you test with Bluetooth, that is part of your number. |
Practical consequence: a 230 ms score on a 144 Hz desktop and a 230 ms score on a budget phone are not the same reflex. Compare scores within one device, and be suspicious of anyone claiming a precise millisecond difference between two people on different hardware.
Getting a number worth keeping
- 1Do a warm-up round, then discard it. First attempts are usually 20–40 ms slow while you figure out the task.
- 2Use the same device and the same time of day for comparisons.
- 3Take the average of five rounds, not your single best. One lucky tap is noise.
- 4Reject rounds where you jumped the gun — the test voids them for you on purpose.
A sudden drop in reaction speed — after a head injury, a new medication, or without an obvious reason — is a matter for a clinician, not for a web test.