How to Improve Your Reaction Time: What Actually Helps

What your reaction time score is made of

A reaction time score is the gap between a cue and your response, and part of it is not you. You notice the cue, decide and move; meanwhile the screen has to draw the cue and the browser has to register your tap or click. Studies that timed browsers with external equipment found that this delay differs between devices, browsers and setups, and no setup was best in every condition (Anwyl-Irvine and colleagues, 2021).

That is why the first rule of improving is to measure fairly. Our comparison checklist goes through what has to stay the same before two scores can be compared.

Practise the exact task you want to improve

Practice reliably makes you faster at the task you practise and at closely related ones. Whether that speed carries over to other situations is far less certain, as our page on the science of brain training explains. So choose the practice that looks like your goal.

1. Pick one task, for example the reaction time test if you want a faster simple reaction.
2. Do short sessions of a few minutes rather than one long one, and stop when your attention drifts.
3. Note the average of each session, not only the best trial.
4. Change one thing at a time (the task, the device, the time of day), so you know what moved the score.

Simple and choice reactions are different skills

In a simple reaction there is one cue and one response: the screen turns green, you tap. In a choice reaction you first have to decide which response is right, and that takes longer. Hick (1952) showed that choice reaction time rises as the number of possible signals grows, roughly in line with the logarithm of the number of choices.

If the speed you care about involves a decision, practise decisions. Reflex Duel asks you to compare two expressions and tap the bigger one, and the goal can switch to SMALLER without warning, so speed only counts when the answer is right.

Wait for the cue instead of guessing it

A common way to get a falsely fast score is to tap on rhythm, before the cue appears. Good tests make this pointless: in the Altamens reaction time test the wait before the colour change is random, between 1.5 and 4 seconds, and an early tap is discarded and the trial restarts.

Keep your finger ready and your eyes on the panel, but react to the change, not to your sense of when it is due. Stopping a response you have prepared is a skill of its own: in Signal Switch a cross means hold still, and tapping on it costs you.

Test when you are rested and alert

Tiredness shows up clearly in reaction tasks. A meta-analysis of 70 studies on short-term total sleep deprivation found the largest effect on lapses in simple attention, which is the kind of task a reaction time test is (Lim and Dinges, 2010).

In practice: compare sessions done at a similar time of day and in a similar state. A slow session after a short night does not mean you have lost what you practised.

Take delay out of your setup

You cannot practise away the delay of a screen or a touchpad, but you can stop it from changing between sessions.

• Use the same device, browser and input method each time: a phone screen, a mouse and a touchpad give different results.
• Close heavy tabs and apps that compete for the device while you test.
• When you change device, start a new baseline instead of comparing with old scores.
• Do not subtract a guessed device delay from your score: without measuring equipment it replaces one unknown with another.

How to tell whether you have really improved

Compare averages across several sessions made under the same conditions. Your best single trial is an extreme: the more attempts you make, the more chances you have of one lucky fast tap, so a new personal best can appear without any real change. Our methods note on reaction time summaries shows with worked numbers how the mean, the median and the fastest trial react differently.

A worked example: on Monday your five trials are 300, 310, 320, 330 and 340 ms, an average of 320 ms with a best of 300 ms. On Thursday they are 280, 330, 340, 350 and 360 ms. The 280 ms tap is a new personal best, yet the average is 332 ms, slower than Monday. One fast trial is not progress; a lower average that repeats is.

A trend that holds over a week or two of sessions is worth more than one good day.

What reaction time practice will not do

A faster score in a browser task shows that you got faster at that task. It does not measure general ability, driving fitness or health, and the Altamens test has no validated population norms. If you notice a sudden change in how quickly you respond in daily life, that is a question for a doctor, not for an online test.