How to train your reaction time (and what actually improves)
CheckReactionTime.org ·
Yes, you can train your reaction time — but far less of it is trainable than most training pages admit, and the part that does move is not the part people expect. Most of the improvement you will see in your first week is you learning the task, not your nervous system getting quicker.
That distinction decides whether training is worth your time. This guide separates the fixed floor from the trainable gap, gives you drills and a four-week plan, and shows you how to measure progress in a way that will not flatter you. Everything below is written to be used alongside the free reaction time test on this site, which is where your baseline comes from.
On this page
- Can you train your reaction time? — whether reaction time is trainable, how much it moves, and how long it takes
- What training actually moves — the fixed biological floor versus the 30 to 60 ms you can actually close
- Reaction time drills that work — reaction time drills that work, what each trains, and how long each takes
- A four-week reaction time training plan — a four-week reaction time training plan at ten minutes a day
- How to measure progress without fooling yourself — median of five runs, fixed hardware, and what score to aim for
- Sleep, caffeine and exercise: what changes your score today — sleep, caffeine and exercise: what shifts your score on the day
- Do reaction time games and apps actually work? — whether reaction time games, apps and action games actually transfer
- Training mistakes that waste weeks — the training mistakes that waste weeks
- Train on all six reaction tests — training across the reaction time, Go / No-Go, aim, CPS and audio tests
Can you train your reaction time?
Reaction time improves with practice, and the effect is real and repeatable. Where it comes from is the interesting part. A simple reaction — see the signal, click — has almost no decision in it, so there is very little for practice to optimise. What improves is everything around the reaction: your expectation of when the signal is coming, your readiness to move, how tight your finger sits on the button, and how little you hesitate about whether the signal was real.
Expect most of the visible change in the first two weeks, and expect it to be mostly familiarity. Someone starting at 280 ms will usually settle somewhere in the 220 to 240 ms range after a few weeks of regular practice, and then stop. Genuine gains in readiness and consistency accumulate more slowly, over four to six weeks, and are smaller than that first-week jump. The plateau after it is not a lack of effort — it is the floor described below, and no amount of training moves it.
What training actually moves
It helps to split your score into two parts, because only one of them responds to training.
The part that is fixed
A large part of your reaction time is transport and chemistry. Light has to reach your retina and be converted into a nerve signal, that signal has to cross several synapses on its way to the cortex, the movement command has to travel back down to your hand, and your muscle has to physically contract. Each of those steps costs time that no practice removes. Together they account for most of a typical visual reaction time.
Age moves this floor too, and it moves it whether you train or not. In a large population study, reaction times were shown to sharpen into the early twenties and then slow gradually across adult life. Training does not reverse that trend, which is why you can train at any age but should expect the range you are training within to shift. An older person who practises regularly may well score the same as a younger person who never has. What training does is keep you at the top of your own range rather than somewhere below it.
The part that is trainable
The trainable part is the gap between your floor and what you actually score. That gap is made of attention, anticipation, readiness and hesitation, and for most people it is worth 30 to 60 ms — which is a large fraction of what separates a mediocre score from a good one.
The catch is that this gain is specific. Practice tends to improve the trained movement under the trained conditions, and the specificity of practice has been shown repeatedly for exactly this kind of goal-directed response. Training on a visual click test makes you better at visual click tests. It does not make you better at everything, and a page telling you otherwise is selling something.
Reaction time drills that work
Drills only help if each one targets a different part of the gap. Doing the same test twenty times trains anticipation and nothing else. The set below covers readiness, inhibition, search and precision, which is what the gap is actually made of.
| Drill | What it trains | How long |
|---|---|---|
| Five clean runs on the reaction time test | Baseline readiness and consistency | 2 minutes |
| Go / No-Go at a high go ratio | Inhibition: the ability not to respond | 3 minutes |
| Aim trainer on small targets | Precision under time pressure | 3 minutes |
| Sequential targets, 15 or 25 count | Visual search while moving | 2 minutes |
| Audio reaction test with wired headphones | The faster sensory channel, and attention without a screen cue | 2 minutes |
Rotate through them rather than grinding one. Five to ten minutes total is enough, and beyond about fifteen minutes fatigue starts working against you and your scores stop meaning anything.
A four-week reaction time training plan
The plan below assumes about ten minutes a day, five days a week. It front-loads familiarity, because the first week is mostly you learning the tasks, and only starts treating your numbers as data in week two.
| Week | Focus | What to do |
|---|---|---|
| 1 | Familiarity | Run each of the six tests once a day. Ignore the numbers entirely; you are learning the tasks, not measuring yourself. |
| 2 | Baseline | Same rotation, but record the median of five runs on the reaction time test. This is your real starting point, not week one. |
| 3 | Load | Add Go / No-Go and the aim trainer at harder settings. Expect your scores to get worse before they settle. |
| 4 | Consistency | Return to week two settings and compare medians. Judge the month on the change in median, never on your best single run. |
If you miss days, repeat the week rather than skipping ahead. The schedule matters more than the total: spreading practice across short frequent sessions beats concentrating it, an effect that holds up in controlled comparisons of practice schedules and that most people get backwards. Ten minutes daily will outperform an hour on Sunday, and the long session is actively worse because fatigue degrades both your scores and the quality of what you are learning.
How to measure progress without fooling yourself
This is where most self-directed training goes wrong. Reaction time is noisy enough that you can convince yourself of an improvement that is not there, simply by picking which numbers to remember.
| Rule | Why it matters |
|---|---|
| Take the median of five runs | Reaction time is noisy. A single run mostly measures luck, and your best run measures it entirely. |
| Test on the same hardware every time | Screen refresh rate and input device shift results by more than a month of training does. |
| Test at the same time of day | Alertness follows a daily cycle, so morning and evening scores are not comparable. |
| Discard false starts, do not average them | A false start is a failed trial, not a fast one. Keeping them corrupts the average in both directions. |
| Compare only with your own history | Someone else's number tells you about their equipment and age as much as their reflexes. |
The single most useful habit is to record the median of five runs rather than your best. Your best score is a measure of your luck; your median is a measure of you. As for what to aim at: on a simple visual test most adults land between 200 and 250 ms, and anything consistently under 200 ms is fast — but rather than chasing a public number, move your own median down and then hold it, because your equipment and your age both shift the scale you are being measured on.
Sleep, caffeine and exercise: what changes your score today
Three things move your score on the day, independent of any training. They are worth knowing because they can easily be larger than a month of practice, and because they contaminate your progress data if you ignore them.
| Factor | Effect on your score | What the evidence says |
|---|---|---|
| Sleep | Large | Short-term sleep deprivation degrades attention and speed measurably, and a meta-analysis found the effect on simple attention tasks to be among the largest. One bad night outweighs a month of practice. |
| Caffeine | Small to moderate | A systematic review of caffeine on reaction time and marksmanship found reliable improvements in speed, with accuracy less consistently affected. Useful, but it moves your score without moving your ability. |
| Exercise | Small, short-lived | A meta-analysis of acute exercise and cognitive performance found a modest benefit shortly after moderate exercise. Enough to matter for a test, not enough to train with. |
| Testing while tired | Large and negative | Fatigue within a single session is visible in your own data: compare your first third with your last third. |
None of these are training. They are the conditions under which you should test, and holding them steady is what makes week-to-week comparison possible at all.
Do reaction time games and apps actually work?
This is the question the category is built on, and the answer splits cleanly in two. Action games do produce measurable differences in visual attention: Shawn Green and Daphne Bavelier showed in 2003 that experienced action gamers track more objects at once and process visual detail faster, and that non-gamers trained on an action game improved on the same measures. That finding is real and has been replicated.
What has not survived is the broader promise built on top of it. A comprehensive meta-analysis of video game training found no reliable transfer to general cognitive ability once study quality was accounted for. The practical reading is the same as everywhere else on this page: train the specific thing you want to be good at. A reaction test trains you for reaction tests, and that is worth doing if a fast response is what you are after — just do not expect it to make you generally sharper.
Training mistakes that waste weeks
The most common mistake is testing to feel good rather than to find out. Screenshotting your best run, retesting until you beat a number, comparing yourself with a stranger on different hardware: none of it tells you anything, and all of it feels like progress.
The second is training tired. A session run at the end of a long day measures your fatigue, not your reaction time, and it teaches you a slower rhythm. If your scores in the last third of a session are consistently worse than the first third, the session is too long.
Train on all six reaction tests
Reaction time is not one skill, and training only the click-on-green test leaves most of it untouched. Each test on this site isolates a different part of the chain, which is what makes them useful as a rotation rather than as six versions of the same thing.
Start with the reaction time test for your baseline, then add Go / No-Go for inhibition, the aim trainer for precision, sequential targets for visual search under time pressure, the audio reaction time test to train the faster of your two senses, and the CPS test for raw finger speed. All six are free and run in the browser.