What a “brain age” score actually measures

Cognitive testing · updated 2026-09-05

If you have taken an online brain-age test, you were handed a single number: your brain is 34, or 52, or 27. It feels precise. It is worth understanding what actually produced it, because the answer changes how much weight the number deserves.

Where the number comes from

Almost every brain-age test is built from two ingredients. The first is accuracy: how many items you got right. The second is speed: how long you took to answer, usually measured as a median across items so that one interruption does not distort the whole run.

Those two numbers get combined into a single score, and that score is then mapped onto an age scale. The mapping is the part people misunderstand. There is no sense in which a score of 0.78 is 34 years old. Someone chose a curve that turns scores into ages, and chose it so that typical performance lands somewhere flattering and plausible.

A brain-age number is a presentation of your score, not a measurement of your brain. The score is real. The age is a costume it is wearing.

Why speed is included at all

Speed is not padding. On tasks that are easy enough that most people eventually get them right, how long you take is often more informative than whether you succeeded. Two people who both answer every item correctly, one averaging 1.2 seconds and one averaging 6 seconds, are not performing identically, and an accuracy-only score cannot tell them apart.

This is also why brain-age scores are sensitive to things that have nothing to do with cognition. Answering on a phone with your thumb is slower than a mouse. A slow connection adds delay. Being interrupted halfway through inflates your median. None of that reflects the thing the test claims to measure, and all of it moves the number.

Why your score moves between attempts

People often retake a test and get a materially different result, then wonder which one was right. Usually neither is wrong; the variance is expected. Several things move it:

• Practice. The second time you see a task type you already know the strategy, which is why most tests warn you that a repeat attempt is not comparable.
• Time of day. Reaction time varies across the day for most people, and the direction is individual.
• Item sampling. A 14-item test is a small sample. Getting two hard items instead of two easy ones is a meaningful swing.
• Effort. Attention is not constant. A run where you were half-watching something else is a genuinely different run.

A short online test has enough measurement noise that differences of a few years should be treated as nothing at all.

What it is good for

Taken as entertainment with a grain of signal, a brain-age test does a few things well. It gives you a reason to pay attention to your own speed and accuracy, which most people never observe directly. It makes the difference between task types visible: many people discover they are far stronger on pattern reasoning than on short-term recall, or the reverse, and that split is more interesting than the headline number.

What it is not good for is anything clinical. A short unsupervised test taken on your own phone, with no controlled conditions and no normative sample, cannot tell you anything about your health. If you are worried about your memory or concentration, that is a conversation for a qualified professional, not a quiz.

Reading your own result sensibly

Three habits make the number more useful:

1. Look at the breakdown, not the headline. Which domains carried your score, and which dragged it down? That is the part that might tell you something.
2. Compare yourself to yourself. A result taken under similar conditions a month apart is more informative than a comparison to a stranger.
3. Ignore small differences. If the number moved by less than about five years, treat it as unchanged.

The best use of a score like this is as a prompt rather than a verdict: something that makes you curious about how your attention and memory actually behave, rather than something that settles the question.