Reaction Time test online: simple and choice

Press a key the moment a circle appears, then press the arrow key that matches an arrow. Your two average times, and the gap between them, show how fast you respond to a signal and how much a choice adds.

Start the Reaction Time test

2 parts · about 4 min · keyboard

After the arrow part you see both of your times and can keep them with a free account.

2 parts · about 4 min · keyboard

This task needs a computer with a physical keyboard. Open this page on a computer to take it.

Start Symbol Coding

Symbol Coding works on a phone: match digits to symbols against a 90-second clock.

Circle: Space. Arrow: arrow keyPart 1 simple · part 2 choice

How it works

5 practice trials, then 2 timed parts: 20 trials with a circle and 40 with an arrow. Every trial lasts only a few seconds, so the whole test takes about 4 minutes.

  1. 1 · First

    5 practice trials

    The circle task with feedback: your time in milliseconds after each press, or a warning if you pressed too soon. Practice is not scored.

  2. 2 · Part 1

    Space for the circle

    A plus sign holds the centre for a random 1 to 4 seconds, then a circle appears. Press Space as fast as you can. With no press after 2 seconds, the trial counts as missed.

  3. 3 · Part 1

    Sometimes nothing comes

    2 extra trials show no circle at all. If you press during that wait, a message asks you to wait for the circle.

  4. 4 · Part 2

    Left or right

    After the plus sign an arrow appears, pointing left or right, 20 times each in a random order. Press the matching arrow key. Speed and accuracy both count.

Only presses after a circle or arrow appears are timed, apart from the 2 waits with no circle, where a press is flagged. The timed parts do not show your time after each press, unlike the practice, and part 2 begins after a short instruction screen with no practice of its own. A bar at the top tracks your progress.

Reaction time was one of the first mental speeds anyone measured. In the 19th century Francis Galton recorded visual reaction times below 190 ms in young people (Woods et al., 2015). Computer versions are far easier to run, but screens, keyboards and software add delays of their own, which is one reason published averages differ from lab to lab.

What it measures

Both parts time the path from seeing a signal to pressing a key. Part 2 adds a decision about which key to press.

Part 1: simple reaction

  • Detection: noticing the circle the instant it appears.
  • Response speed: starting and completing the key press.
  • Alertness: staying ready through waits of different lengths, so the moment cannot be predicted.

Part 2: choice reaction

  • Everything part 1 asks for, with a new signal.
  • A decision on every trial: which way does the arrow point?
  • Accuracy under time pressure, since a fast press on the wrong key is an error.

Subtracting your part 1 average from your part 2 average gives a rough estimate of how long the choice itself takes. The result screen calls this decision time.

Age affects the two parts differently. In a reanalysis of 7,130 adults from a UK health survey, simple reaction time changed little until around 50, while choice reaction time slowed across the whole adult range (Der and Deary, 2006). With precisely timed equipment, Woods and colleagues found that most of the slowing with age came from slower movement rather than slower detection (Woods et al., 2015).

A test, not a diagnosis.

Reaction times from one sitting describe that sitting. They cannot confirm or rule out a diagnosis.

What you get

Your times appear once the last arrow trial is over.

On the result screen

  • Your average simple and choice reaction times in milliseconds, each on a bar from faster to slower with a marker for you.
  • Decision time, choice minus simple: labelled very efficient under 120 ms, normal under 180 ms and elevated from 180 ms.
  • A verdict from fixed cut-offs on both averages, the same at every age.
  • For each part, the median, the spread and the number of valid trials, plus accuracy and errors for the arrows.
  • A CSV file with every trial.

When you keep it

  • Both times are stored with a free account, ready to reopen on any device.
  • They join your processing speed results, beside Symbol Coding.
  • A next test comes up as soon as you finish, so your profile grows one area at a time.

How this version is scored

Each trial stores what appeared, which key you pressed and when. The averages use only the trials that pass the checks below.

Reaction time
The time from the circle or arrow appearing to your key press, in milliseconds.
Valid trial
A press from 150 to 1,500 ms after the signal. Faster presses are treated as anticipation and slower ones as lapses, and both are left out. In part 2 the press must also be on the correct arrow key.
Averages
The mean of your valid trials in each part. The data panel adds the median and the standard deviation, which shows how much your times varied.
Decision time
Your part 2 average minus your part 1 average, in milliseconds.
Enough data
Each part needs at least half of its trials valid: 10 of 20 with the circle and 20 of 40 with arrows. Below that, the screen says the attempt cannot be scored and shows your counts instead of a verdict.
Choice accuracy
Correct arrow keys as a share of the arrow trials you answered. Missed trials are not part of it.

Why the waits vary. If the circle always came after the same pause, you could press on rhythm instead of reacting. A random wait of 1 to 4 seconds, plus 2 trials with no circle, keeps each press tied to the signal.

Your computer is part of the time. Keyboards, screens and browsers each add a few milliseconds, and the amount differs between machines. When Woods and colleagues removed the delays of their own hardware, mean simple reaction time fell from 231 ms to 213 ms (Woods et al., 2015). Web studies can match lab results on many tasks (Germine et al., 2012), but only when you compare like with like, so retake this test on the same computer.

Before you start

Reaction times move with alertness, so a few minutes of preparation pay off.

For a cleaner attempt

  • Use a computer with a physical keyboard; rest a finger on Space in part 1 and on the arrow keys in part 2.
  • Close other tabs and programs that could make the browser stutter.
  • Choose a time when nobody will interrupt you for about 4 minutes.
  • Wait for the signal: pressing early only throws trials away.

Interpret cautiously when

  • You slept badly the night before; short sleep affects attention tasks like this one (Lim and Dinges, 2010).
  • You use assistive input or cannot make quick movements; then the times do not describe standard reaction speed.
  • The computer was slow, a wireless keyboard lagged, or other programs were busy.
  • You were unwell, in pain or taking medication that affects alertness.

Questions

What is a normal reaction time?

It depends on the task and the equipment. In one large study with carefully calibrated hardware, adults aged 18 to 65 averaged 231 ms for simple reaction time, and 213 ms once hardware delays were removed (Woods et al., 2015). A browser test adds delays of its own, so expect somewhat longer times here. The result screen reads your averages against its own fixed cut-offs, not age norms.

What is the difference between simple and choice reaction time?

Simple reaction time has one signal and one response: see the circle, press Space. Choice reaction time has two possible signals, each with its own key, so you have to decide before you press. The extra time reflects that decision, and choice times also slow more with age than simple ones.

Can I take the Reaction Time test on my phone?

No. It needs a physical keyboard, with Space for the circle and the arrow keys for the arrows, so the tests hub asks for a computer. On a phone you can start with Symbol Coding, the other processing speed test, and take this one later on a computer.

Why were some of my trials not counted?

Presses faster than 150 ms after the signal are treated as anticipation, and presses after 1.5 seconds as lapses. Both are left out, as are wrong arrow keys in part 2. The data panel shows how many trials stayed valid in each part, and if fewer than half did, the screen asks you to retake the test rather than giving a verdict.

Sources

By Cristian Dominguez Rein-Loring, Creative Technology Lead · Updated October 12, 2026