The Age Curve, in a Sentence
"Reaction time slows with age" is true, but the picture is different from what people imagine. It's not a cliff — it's a gentle downslope. Looking at simple reaction time, performance peaks in the early 20s and then slows very gradually. The decline is mild enough that even into the 60s and 70s, averages can stay in the 300–400ms range. It's not the dramatic curve of "it's all over after your 20s."
This point matters because talk about age and cognitive speed is often exaggerated or categorical. Looking at the data calmly, the age effect is real but smaller than expected, and individual differences and practice can substantially offset it.
Reaction Time by the Numbers
The numbers vary by study, but the general trend looks like this. The difference in simple reaction time between people in their 20s and people in their 60s is often reported at around 20–50ms. For instance, if the 20s average is 250ms, the 60s average might be around 280–300ms. Of course, this range shifts depending on the task (simple vs. choice reaction), participants' health, and how much they've practiced.
What matters is that it's a gradual decline, not an acceleration. There's no sudden drop between your 30s and 40s. From the 20s onward, reaction time slows a little, at a nearly constant rate, every decade. That's why many people maintain a reaction time well into their 70s that causes no trouble in daily life.
Memory and Computation Age Too
It's not just reaction time. Working memory capacity — the ability to hold information in mind while you work with it — also declines gently with age. On tasks like hearing a string of digits and repeating them backward, accuracy and processing speed tend to dip slightly as age increases.
But again, this is usually not "to a degree that interferes with daily life." Decline within the range of normal aging generally doesn't cause problems with everyday cognitive tasks. Measuring yourself with the Memory game can serve as a reference to compare against age-group averages. Just don't conclude from a single score that "my memory is aging."
Practice Can Offset It
Here's the key. The age effect is path-dependent. In other words, at the same age, outcomes differ depending on how the brain has been used over the years. People who regularly engage in some cognitive task — an instrument, a foreign language, puzzles, games, anything — often maintain strong performance regardless of age. This isn't the exaggeration that "training your brain makes you younger"; it's the plain fact that proficiency at a specific task raises performance on that task.
Reaction time is the same. Repeat the Reaction Time game for a few days and you'll likely see your own numbers get faster than your first session. That's not your brain getting younger — it's the result of becoming familiar with (learning) the procedure. This learning effect shows up clearly even at older ages. So when you look at "age-group averages," remember that those averages mix people who've practiced a lot with people who haven't.
Don't Conclude "Aging" From a Single Score
This is the point I most want to emphasize. A single slow score on Reaction Time or Memory does not mean "my brain is aging." That number is the sum of that day's sleep, the time of day, your condition, your familiarity with the task, and even your mood right before the test — countless factors combined.
To talk about an aging effect, you need a trend from repeated measurements under the same conditions over a long period. A single value is "today's state," not "your brain's age." A more meaningful approach is to take the average of measurements over several days at the same time of day as your personal baseline, and watch how that baseline moves over months.
Also, measurements on this site are not a medical diagnosis and don't replace clinical evaluations for dementia or mild cognitive impairment. If you have symptoms that suggest cognitive decline, consult a neurologist. Keep in mind that the values measured here are for reference — a way to observe your own cognitive speed. The data itself doesn't reduce cognition to the single variable of age, and neither should you.