How to Improve Reaction Time: What Actually Changes It, and by How Much

How to improve reaction time: expect small gains, not big ones. In a ruler drop study of 43 young adults, ten sessions over five weeks cut the average by almost 13 ms (Del Rossi and colleagues, 2014). A good night's sleep and caffeine also help a little. But the device you test on moves a score by more than any habit: adult studies range from 233 to 397 ms.

Below: what practice, sleep, caffeine and video games have been shown to change, in milliseconds where the studies give them, and how to tell a real change from noise.

How to improve reaction time: can practice make you faster?

Yes, but only a little. The best direct evidence we found is a 2014 study in the Journal of Athletic Training. Forty-three people, with an average age of 21.8, did a ruler drop test in 10 sessions over 5 weeks, with 10 drops per session. Their reaction time fell by almost 7 ms between the first two sessions and by almost 13 ms between the first and the tenth.

The change from the first session to most later sessions was statistically significant. The first-to-second and first-to-third comparisons alone were not, and successive sessions did not differ from each other. The authors advise at least one practice session before recording a baseline.

How big is 13 ms? Near the 20 cm mark on a ruler, one centimetre is about 5 ms, so 13 ms is about 2.5 cm. That is our arithmetic with t = √(2d/g); our ruler drop test chart shows the full table. It is far less than the minimal detectable change reported for a single ruler drop, 10.75 cm, which is about 48 ms from a 20 cm catch.

Practice helps on timed tests in general. Our guide to the practice effect on cognitive ability tests reports that a second sitting of an employment test averages about a quarter of a standard deviation higher.

What slows reaction time the most?

Sleep loss slows it the most of anything we found. Lim and Dinges (Psychological Bulletin, 2010) pooled 70 articles with 147 cognitive tests on short-term total sleep deprivation, meaning under 48 hours. The largest effect was on lapses in simple attention, the moments when you fail to respond in time: g = -0.776, with a 95% interval of -0.96 to -0.60. Reasoning accuracy hardly moved: g = -0.125, not significant.

Here g is an effect size in standard deviation units, and by the usual rule of thumb 0.8 counts as large. Lapses are not the same as a slower average click, so read it as: a bad night makes missed or very late responses much more likely. Our guide to sleep loss and test performance goes further.

Age slows reaction time too, but slowly. In a calibrated study of 1,469 adults, simple reaction time rose 0.55 ms a year, about 22 ms between ages 18 and 65, while detection time stayed near 131 ms at every age. Equipment matters far more. Studies of adults in the same age range report anything from 233 to 397 ms, a spread of 164 ms, as our guide to average reaction time by age shows.

Does caffeine improve reaction time?

Does caffeine improve reaction time? Yes, slightly, in rested healthy adults. A 2025 meta-analysis in Psychopharmacology (Kløve and Petersen) pooled 31 randomised, placebo-controlled trials with 1,455 participants. It found an effect of g = 0.28 on reaction time and 0.27 on accuracy.

Doses of 200 mg or more improved reaction time more than lower doses, and the effect did not depend on how much caffeine people usually drank. We give no dosing advice. For scale only: on a test where scores usually spread by 25 ms, 0.28 of a standard deviation is about 7 ms. That 25 ms is the spread in a large athlete sample on our ruler page, used here only as an example, not a result of the review.

Our guide to caffeine and cognitive test performance covers the debate about withdrawal. The meta-analysis above found no link with habitual intake, in people who were rested.

Do video games, exercise or reaction time training make you faster?

We found no pooled result showing that they speed up simple reaction time. We searched PubMed on 14 November 2026 and found no review of reaction time training or exercise that we could read on this outcome. People who search how to increase reaction time usually want a drill, but the evidence we can show points to sleep and practice first.

Action video games have the best-studied claim, and it is about something nearby. A 2018 meta-analysis in Psychological Bulletin (Bediou and colleagues) found that action video game training improved cognition by g = 0.34 on average, with the clearest gains in top-down attention and spatial cognition. It did not single out simple reaction time. The authors also estimate that published effects are about 30% larger than the full literature, because of publication bias.

So treat faster reflexes from training as possible but small. A ruler or a click test measures a simpler chain than an action game trains, and we found no study that links the two.

How do you measure change in reaction time honestly?

Use one method, one device and several trials. A single ruler drop has a reliability (ICC) of 0.57 and a minimal detectable change of 10.75 cm, about 48 ms near 20 cm, in a 2024 study of 51 older adults. Online tests add device lag: Human Benchmark says your computer could add 10 to 50 ms. Both are bigger than the 7 to 13 ms effects above.

• Test on the same device, with the same hand and the same screen setting every time
• Do a practice round first, as the ruler study advises, before you record a baseline
• Take at least five trials and use the middle value, the median
• Test at a similar time of day and after similar sleep
• Compare your own results over weeks, not with someone else's number

You can track your own numbers with our reaction time test guide and the Stroop test game, which adds a decision to the reaction.

Does a faster reaction time raise your IQ?

No. Simple reaction time correlates about -.31 with IQ in 900 adults aged about 56, which is under 10% of the variance, and an IQ cannot be read from a click. Making a click faster will not move an IQ score. The numbers are in our guide to average reaction time and IQ, and our IQ percentile calculator turns any IQ-scale score into a percentile.

To see how your reasoning compares with other adults, take the IQ Metrics test.

Common questions