← All guides

Age and reaction time: what a large survey found, and what it cannot tell you

Everyone has heard that reaction time slows with age. The interesting part is how that was measured, and how little it says about any one person.

Updated November 2, 2026

If you are 55 and just scored slower than a 20-year-old friend, the first thing to know is that the gap is real on average and almost useless for you personally. Averages across thousands of people describe a population. Your own number describes your screen, your sleep and this morning. This guide covers what the best-known survey of the question did, what it found, and how to read your own result next to it.

What the survey did

Geoff Der and Ian Deary published the study in 2006, using the United Kingdom Health and Lifestyle Survey. It measured reaction time in about 7,000 adults across the adult age range, on two tasks. One was simple reaction time: respond when a signal appears, nothing to decide. The other was four-choice reaction time, where each of four signals had its own response, the same ingredient the choice reaction test here adds, and the one the Hick guide explains.

The analysis is cross-sectional. That shapes everything below: it compares different people of different ages measured at one point in time, not the same people as they grow older.

What it found

Three results are the ones worth carrying away.

What a single survey cannot say

Because the comparison is between people, not within them, a difference between a 25-year-old and a 65-year-old in this kind of data mixes two things: the effect of being 40 years older, and everything else that differs between people born in different decades, such as schooling, health history and familiarity with devices. A study that follows the same people over years separates them. This one, by design, does not.

It also measured reaction time on its own apparatus, in its own conditions. The paper’s numbers are the paper’s: a browser test on a phone adds its own delays, so you cannot place your result on their axis by comparing milliseconds. The honest use of the survey is its shape, not its figures.

What this means for your result

humanlag does not ask your age and does not split its curve by it. The reaction time test puts everyone in one pool, which means your percentile answers “how do you compare with everyone who has played here”, not “with people your age”. The Method page declares how that curve is built and how much of it is still an assumption.

For the path a signal travels from eye to finger, and where the milliseconds go, the science of reaction time is the companion read.

Sources

  • Der, G., & Deary, I. J. (2006). "Age and sex differences in reaction time in adulthood: results from the United Kingdom Health and Lifestyle Survey". Psychology and Aging, 21(1), 62–73.

Keep reading

  • What actually makes you fasterMost of what the internet sells for reaction time does nothing. Most of what works is free, and you already know what it is.
  • How much a head can holdThe buffer is tiny, the range is narrow, and the only real upgrade ever found is cheating in a way psychologists approve of.