How to improve reaction time: what the evidence supports
How to improve reaction time? The clearest evidence is about what slows you down, mainly sleep loss, rather than a proven way to get faster. Practice sped up choice tasks and gave mixed results on a plain simple reaction time. Caffeine and a bout of exercise add small effects. If you want to know how to get better reaction time on this site's reaction time test, the practical route is to measure: set a baseline, change one thing, and retest on the same input.
| Factor | What the studies found | How firm it is |
|---|---|---|
| Sleep loss | Slower simple-attention reaction time after 24 to 48 hours awake (g = −0.732).1 | Pooled lab studies of healthy adults. A standardized effect, not milliseconds. |
| Practice | Faster on choice tasks.7 Mixed on simple reaction time: lower after 3 weeks of key-press practice in one pair of studies,28 unchanged in two others.93 | A handful of small studies, mostly read in the abstract only. |
| Caffeine | A small speed-up in rested adults (g = 0.28).4 A larger one after sleep loss (g = 1.11).10 | Meta-analyses of attention tasks, not simple reaction time alone. |
| Action video games | Players responded about 11% faster.5 The training effect is smaller, and published effects may be inflated.6 | Players are self-selected. The reviews come from the group that proposed the effect. |
| One bout of exercise | A small effect, mostly on speed (g = 0.14).11 | Overall cognition, not simple reaction time alone. |
| Alcohol | Reaction time slowed in a regular way across tasks.12 | Controlled lab data. This site gives no guidance on drinking. |
Where a number is a Hedges g or a Cohen's d, it is a standardized difference between groups or conditions, not milliseconds. When we call an effect small or large, that is the authors' own wording.
Reaction time training and exercises: do they work?
Practice is the factor most directly tied to reaction time training, with a catch: it depends on the task. In a 10,000-trial lexical decision task, which is a choice task, mean response times decreased, and practice changed more than speed. It also changed response caution, response bias and peripheral processing time.7
In 16 male students who practiced a key-press reaction time task, with 3 blocks of 25 trials, 5 days a week, for 3 weeks, reaction time decreased in the peripheral and the central visual field.2 The gain was still there 3 weeks after practice stopped.8 The follow-up paper's title calls this simple reaction time. The abstracts report the direction, not milliseconds, and the group is small.
Other studies found no change in a plain simple reaction time. In 94 high school students followed for 2 years, hitting practice improved Go/Nogo reaction time while simple reaction time stayed constant.9 In 45 adults who repeated a 3-minute vigilance test 16 times, mean and median reaction time did not change systematically. The fastest 10% of reaction times got slightly faster and false starts fell, and the authors call the effects small.3
A longer run is not training. In one session of 120 simple reaction time trials, reaction time rose from 228.1 ms in the first block of 20 trials to 237.2 ms in the last, a small fatigue effect in 1,469 adults.13 That is one reason the test here is five rounds.
Reaction time exercises sold online, such as ruler drops, light boards and apps, are a different matter. We did not find a source we could read that compares named drills, so this page does not rank them. What the studies support is narrower: if you practice a task, expect to get better at that task. The ruler drop test is a way to measure, not a proven drill.
What a browser test can show you
A browser test can show whether your own median moves between sessions on the same setup. It cannot show why it moved, and it cannot separate you from your equipment.
Some of your number is not you. The signal can only appear on a screen refresh, so you wait up to one frame: 16.7 ms at 60 Hz, 6.9 ms at 144 Hz. In one lab study the monitor added 11.0 ms and a gaming mouse 6.8 ms, and standard mouse drivers can add 20 ms or more.13 A change of screen or mouse can move your number without any change in you, which is why a retest keeps the same setup.
How to retest fairly
- Run it once now. Your first finished run is the baseline. A false start or a guess under 100 ms does not count.
- Change one thing. A full night of sleep, a set number of practice days, or nothing at all.
- Retest on another day. Use the same input, device, browser and screen setting, at a similar time of day, rested.
- Compare medians. The result card shows the median of five rounds, which a single slow round does not move much.
- Repeat on several days before you read anything into it. Day-to-day variation is normal.
The average reaction time page lists published figures with how each was measured, and the page on what reaction time is explains the parts of a response.
Read how this test measures for the sources of error and the rules behind each round.
How to improve reaction time FAQ
How can you improve reaction time?
The studies point to a short list for how to improve your reaction time. Sleep loss slows reaction time, so being short of sleep is the clearest thing to avoid. Practice sped up choice tasks, and its effect on a plain simple reaction time was mixed: lower after three weeks of key-press practice in one study, unchanged in two others. Caffeine and a bout of exercise add small effects. No study we read shows a guaranteed way to get faster, so measure your own baseline and retest instead of expecting a result.
Can reaction time be improved with practice?
It can, depending on the task. In a 10,000-trial choice task mean response times decreased. In a 3-week study of key-press practice reaction time fell and the gain was still there 3 weeks later. Two other studies found no change in a plain simple reaction time: 45 adults who repeated a 3-minute vigilance test 16 times showed no systematic change in mean or median reaction time, and 2 years of baseball hitting practice left simple reaction time constant. Practice effects depend on the task you practice.
How to increase reaction time, and what does faster mean?
Reaction time is a duration, so increasing it makes you slower. Most people who search this, or how to get faster reaction time, mean the opposite: a shorter reaction time, which is a faster response. On this page, "faster" means a lower number of milliseconds. The factors that lengthen it, such as sleep loss, are covered above so you can avoid them.
Is a reaction time trainer worth it?
We found no study comparing named reaction time trainers, so this page cannot rank one. The evidence we did read says practice effects depend on the task: a trainer teaches the task it trains, and the gain may not carry over to a different one. If you try one, set a baseline with the reaction time test first and retest on another day with the same setup.
What are the best exercises for reaction time?
We did not find a source we could read that compares named reaction time exercises such as ruler drops, light boards or apps, so this page does not rank them. The studies we did read test practice of one task, action video games, and a single bout of aerobic exercise. Those effects are small or specific to the task, as summarized above.
How to train reaction time without fooling yourself?
Set a baseline with the reaction time test, practice or change one thing, then retest on another day with the same input, device, browser and screen setting. Compare medians, not single rounds. In one practice study, people practiced 3 blocks of 25 trials, 5 days a week, for 3 weeks, so a retest after a few weeks fits that design. This is a browser game, so treat the result as a comparison with yourself.
Does caffeine improve reaction time?
In rested healthy adults, a meta-analysis of 31 randomized placebo trials found a small speed-up (g = 0.28), and a different analysis found a larger effect after sleep loss. Caffeine is a drug, and this page does not recommend taking it or how much.
Why is my reaction time slower than last time?
Day-to-day variation is normal. Tiredness, a different input device, a different screen refresh rate, a distraction or a short wait before the signal can all move your median. A browser reaction time is you plus your screen, input device and browser. Run it on several days before reading anything into one result.
Sleep and reaction time
A meta-analysis pooled 70 articles and 147 cognitive tests in healthy adults aged 18 and over, kept awake for no more than 48 hours. Simple-attention reaction time was slower, with Hedges g = −0.732 (95% CI −0.874 to −0.590; 26 effect estimates), which the authors describe as a moderate to large effect.1 That is a standardized effect. It does not say how many milliseconds a missed night costs, and this page does not give one.
Less than a full night of lost sleep also appears in the data. In 48 healthy adults aged 21 to 38 who spent 4, 6 or 8 hours in bed for 14 days, the authors wrote that sleeping 6 hours or less a night produced cognitive performance deficits equal to up to 2 nights of total sleep loss. That result is about cognitive performance in general, not reaction time alone.14
For your retest, this means a short night is a reason for a different number, not a training result. Retest rested. These are lab results in healthy adults, not advice for any person.
Caffeine and reaction time
A meta-analysis of 31 randomized, double-blind, placebo-controlled trials with 1,455 participants found small effects in favor of caffeine: g = 0.28 for reaction time and g = 0.27 for accuracy, larger with higher doses and unrelated to how much caffeine people usually had. The reaction time there is a composite across attention tasks, not a simple reaction time alone.4
After sleep loss the effect was larger: reaction time g = 1.11 (95% CI 0.75 to 1.47; 12 effect estimates) in a separate meta-analysis.10
One controlled study of 24 first-person shooter players measured a vigilance-test reaction time of 290.8 ms after water, 279.8 ms after a low caffeine dose and 273.0 ms after a higher one. The two doses did not differ significantly. It is a small study, single-blinded, with water as the control rather than a taste-matched placebo.15
Caffeine is a drug. This page reports what studies measured and does not say whether or how much to take.
Action video games and reaction time
A review that pooled 7 studies across 89 conditions found habitual action game players responding about 11% faster than non-players on speeded choice tasks, with effect sizes (Cohen's d) from 0.48 to 1.47 depending on the task, and the same accuracy in both groups. It is the authors' own pooling, not a formal meta-analysis.5
The same review describes a training study in which 25 non-players were randomly assigned to an action game or a control game for 50 hours over 8 to 9 weeks. Reaction time fell 13% with the action game and 6% with the control game. That figure is credited to a lab thesis, so it is second-hand, and the 6% drop in the control group is a practice and retest figure.5
A later meta-analysis found g = 0.55 when comparing players with non-players and a smaller g = 0.34 in training studies, and estimated that published effects are 30% larger than in the full literature. It covers perception, attention and spatial skills, and does not isolate reaction time.6
Read this with care. Players choose to play, the reviews come from the group that proposed the effect, and the training evidence comes from small trials. Nothing here says that playing a game will make your reaction time faster.
Exercise and reaction time
Across 79 studies, one bout of exercise had a small positive effect on cognitive performance overall (g = 0.097).16 A second meta-analysis found a small effect (g = 0.14) that was mostly on speed rather than accuracy, and larger for moderate intensity than for low or high.11
Timing mattered in a third: performance was slightly worse during the first 20 minutes of exercise (g = −0.14) and better afterward (g = 0.20).17 None of the three isolates simple reaction time, so the fair summary is small, and mostly speed.
Alcohol and reaction time
In a meta-analysis that pooled 25 reaction time task conditions, doses of 0.8 or 1.0 ml per kg of body weight slowed reaction time in a regular way: a linear fit accounted for 99.7 per cent of the variance.12
In 39 people, staying awake 17 to 19 hours gave performance on some tests equal to or worse than at a blood alcohol level of 0.05%. Response speeds were up to 50% slower on some tests, which is the upper end, not a general result.18 This site gives no guidance on drinking or driving.
Age, waiting time and equipment
Some things move your number and cannot be trained. In 1,469 adults, simple reaction time rose by 0.55 ms per year of age, mostly through slower motor output, and was unaffected by sex or education.13 The wait before the signal matters too: the shortest wait in that study gave a reaction time longer by roughly 15%.13 The reaction time test here uses a random wait of 1.5 to 4.5 seconds.
Sources
- Lim, Dinges. A meta-analysis of the impact of short-term sleep deprivation on cognitive variables. Psychological Bulletin 136(3):375-389 (2010). doi:10.1037/a0018883. https://doi.org/10.1037/a0018883 (read: full text, checked Nov 8, 2026)
- Ando, Kida, Oda. Practice effects on reaction time for peripheral and central visual fields. Perceptual and Motor Skills 95(3):747-751 (2002). doi:10.2466/pms.2002.95.3.747. https://doi.org/10.2466/pms.2002.95.3.747 (read: abstract only, checked Nov 8, 2026)
- Basner, Hermosillo, Nasrini, McGuire, Saxena, Moore, Gur, Dinges. Repeated administration effects on Psychomotor Vigilance Test performance. Sleep 41(1):zsx187 (2018). doi:10.1093/sleep/zsx187. https://doi.org/10.1093/sleep/zsx187 (read: abstract only, checked Nov 8, 2026)
- Kløve, Petersen. A systematic review and meta-analysis of the acute effect of caffeine on attention. Psychopharmacology 242(9):1909-1930 (2025). doi:10.1007/s00213-025-06775-1. https://doi.org/10.1007/s00213-025-06775-1 (read: abstract only, checked Nov 8, 2026)
- Dye, Green, Bavelier. Increasing speed of processing with action video games. Current Directions in Psychological Science 18(6):321-326 (2009). doi:10.1111/j.1467-8721.2009.01660.x. https://doi.org/10.1111/j.1467-8721.2009.01660.x (read: full text, checked Nov 8, 2026)
- Bediou, Adams, Mayer, Tipton, Green, Bavelier. Meta-analysis of action video game impact on perceptual, attentional, and cognitive skills. Psychological Bulletin 144(1):77-110 (2018). doi:10.1037/bul0000130. https://doi.org/10.1037/bul0000130 (read: abstract only, checked Nov 8, 2026)
- Dutilh, Vandekerckhove, Tuerlinckx, Wagenmakers. A diffusion model decomposition of the practice effect. Psychonomic Bulletin & Review 16(6):1026-1036 (2009). doi:10.3758/16.6.1026. https://doi.org/10.3758/16.6.1026 (read: abstract only, checked Nov 8, 2026)
- Ando, Kida, Oda. Retention of practice effects on simple reaction time for peripheral and central visual fields. Perceptual and Motor Skills 98(3):897-900 (2004). doi:10.2466/pms.98.3.897-900. https://doi.org/10.2466/pms.98.3.897-900 (read: abstract only, checked Nov 8, 2026)
- Kida, Oda, Matsumura. Intensive baseball practice improves the Go/Nogo reaction time, but not the simple reaction time. Cognitive Brain Research 22(2):257-264 (2005). doi:10.1016/j.cogbrainres.2004.09.003. https://doi.org/10.1016/j.cogbrainres.2004.09.003 (read: abstract only, checked Nov 8, 2026)
- Irwin, Khalesi, Desbrow, McCartney. Effects of acute caffeine consumption following sleep loss on cognitive, physical, occupational and driving performance: a systematic review and meta-analysis. Neuroscience & Biobehavioral Reviews 108:877-888 (2020). doi:10.1016/j.neubiorev.2019.12.008. https://doi.org/10.1016/j.neubiorev.2019.12.008 (read: abstract only, checked Nov 8, 2026)
- McMorris, Hale. Differential effects of differing intensities of acute exercise on speed and accuracy of cognition: a meta-analytical investigation. Brain and Cognition 80(3):338-351 (2012). doi:10.1016/j.bandc.2012.09.001. https://doi.org/10.1016/j.bandc.2012.09.001 (read: abstract only, checked Nov 8, 2026)
- Maylor, Rabbitt. Alcohol, reaction time and memory: a meta-analysis. British Journal of Psychology 84(3):301-317 (1993). doi:10.1111/j.2044-8295.1993.tb02485.x. https://doi.org/10.1111/j.2044-8295.1993.tb02485.x (read: abstract only, checked Nov 8, 2026)
- Woods, Wyma, Yund, Herron, Reed. Factors influencing the latency of simple reaction time. Frontiers in Human Neuroscience 9:131 (2015). doi:10.3389/fnhum.2015.00131. https://doi.org/10.3389/fnhum.2015.00131 (read: full text, checked Nov 8, 2026)
- Van Dongen, Maislin, Mullington, Dinges. The cumulative cost of additional wakefulness: dose-response effects on neurobehavioral functions and sleep physiology from chronic sleep restriction and total sleep deprivation. Sleep 26(2):117-126 (2003). doi:10.1093/sleep/26.2.117. https://doi.org/10.1093/sleep/26.2.117 (read: abstract only, checked Nov 8, 2026)
- Rogers, Trotter, Johnson, Desbrow, King. Caffeine improves the shooting performance and reaction time of first-person shooter esports players: a dose-response study. Frontiers in Sports and Active Living 6:1437700 (2024). doi:10.3389/fspor.2024.1437700. https://doi.org/10.3389/fspor.2024.1437700 (read: full text, checked Nov 8, 2026)
- Chang, Labban, Gapin, Etnier. The effects of acute exercise on cognitive performance: a meta-analysis. Brain Research 1453:87-101 (2012). doi:10.1016/j.brainres.2012.02.068. https://doi.org/10.1016/j.brainres.2012.02.068 (read: abstract only, checked Nov 8, 2026)
- Lambourne, Tomporowski. The effect of exercise-induced arousal on cognitive task performance: a meta-regression analysis. Brain Research 1341:12-24 (2010). doi:10.1016/j.brainres.2010.03.091. https://doi.org/10.1016/j.brainres.2010.03.091 (read: abstract only, checked Nov 8, 2026)
- Williamson, Feyer. Moderate sleep deprivation produces impairments in cognitive and motor performance equivalent to legally prescribed levels of alcohol intoxication. Occupational and Environmental Medicine 57(10):649-655 (2000). doi:10.1136/oem.57.10.649. https://doi.org/10.1136/oem.57.10.649 (read: abstract only, checked Nov 8, 2026)
Each source was found by its DOI and read on Nov 8, 2026, either in full or, where marked, in the abstract only; what we say about an abstract-only source goes no further than the abstract. Effect sizes are as the authors report them.