The Brain Is Not "Finished at 25"

This is the most repeated sentence about adolescence in the English language, and it is half right. The half that is right matters, so this page will not throw it out. The half that is wrong is the number, and the number is the part people act on.

What does the federal source actually say?

The National Institute of Mental Health publishes a short public document called The Teen Brain: 7 Things to Know, revised in 2023. It is the source most often waved at when someone says the brain is finished at 25.

Its actual words are: “The brain finishes developing and maturing in the mid-to-late 20s.” And: “The part of the brain behind the forehead, called the prefrontal cortex, is one of the last parts to mature.”

Notice what is missing. NIMH does not say 25. It gives a range, and the range is deliberate. Nothing in that publication attaches a finding to the number that ended up in circulation.

Why is a range instead of an age?

Because the people who study maturation do not think a single age exists to be found.

The neuroscientist Somerville published a paper in Neuron in 2016 asking what we are actually searching for when we search for brain maturity. Her argument, as quoted in coverage of that paper: “brain maturation does not map on to a single developmental timeline, and it differs by individual when different parts of the brain develop.” On the appeal of a single number, she said: “The very idea that we could come up with some number that would encompass all of the complexity involved in brain development is a challenge.”

Two specific things from that work are worth carrying around.

First, on individual variation: some eight year old brains showed greater connectivity maturation on some measures than some twenty-five year old brains. The spread between individuals is wider than the difference the age is supposed to mark.

Second, on the finish line: several regions of the brain had not plateaued even by the age of 30. If you were looking for the moment development stops, thirty is a poor candidate and twenty-five is a worse one.

Somerville also said something that applies directly to the legal ages we already live with: “The age of 18 doesn’t have any biological magic to it.” Neither does 25. Both are administrative lines that got dressed in biology afterwards.

What is actually supported?

Three things, and they are worth having straight.

• The prefrontal cortex is among the last regions to reach an adult-like structure.
• Development continues well past 18, into the mid to late twenties on NIMH’s own account.
• Timing varies enormously between individuals and between regions of one brain.

What is not supported is that 25 is a threshold, that anything completes or switches on at that age, that the number has any biological edge, or that a 24 year old and a 26 year old differ in some categorical way.

It is a slope. It is not a cliff.

Why does this matter for a parent tonight?

Because of how the sentence gets used.

“His brain isn’t done until 25” arrives in a house at a very particular moment: after something has been broken, or shouted, or reported by a school. It is offered kindly, usually by someone trying to take the sting out of it. And it does two things at once, both unhelpful.

It tells you the behaviour was inevitable, which is not what the evidence says and takes the teenager out of the picture entirely. And it quietly tells you that holding a line is unfair, which leaves you standing there at eleven at night with a consequence half applied and no confidence in it.

A slower-maturing brain is an argument for repetition, not for removal. It predicts that the learning takes longer and needs more goes at it. It does not predict that the learning cannot happen, and it says nothing at all about whether a particular sixteen year old should replace a door.

The thing that does move the needle is not waiting out a birthday. In the largest meta-analysis of anger treatment in young people, covering 40 studies and 1,953 children and adolescents, the ingredients associated with the bigger effects were being shown a skill, practising it, and being told how you did. Time on its own was not one of them. Does anger management actually work sets out what those effects actually amount to.

What this looks like in an ordinary house

Gideon is sixteen and has just been suspended for two days after a shoving match in a corridor. This is an illustration rather than a real family.

His mother says the sentence to herself in the car park, twice. His brain is not done yet. It is not really his fault. By the time she gets home she has talked herself out of the consequence she had decided on that morning, because applying it now feels like punishing a child for his own biology.

Nothing about the neuroscience required that. What the evidence supports is that his judgement is still being built, slowly, out of exactly the sort of ordinary friction she was about to remove.

She applies the consequence she had planned, which is dull and small and specific. She also drops the speech about how he could have ruined his future, because that was doing no work. Then she asks the school a narrow question about what happened in the ten minutes beforehand. Suspended for fighting covers what those first forty-eight hours involve.

Six months later Gideon is not a different person and his brain is no more finished than it was. He has had four arguments that did not become incidents, which is the actual unit of progress here.

Is there a version of this I can say out loud?

Yes, and it is not much longer than the wrong one.

The part of the brain behind the forehead is one of the last to mature, and maturing runs into the twenties and for some regions beyond. There is no switch at 25. Brains vary so much from person to person that age is a rough guide at best.

That version is true, it is short enough to say in a kitchen, and it does not hand a fifteen year old a doctor’s note for the next four years.