The brain's neurons, counted
From the 2009 count and the review you'll find in the sources below.
Where the 86 billion comes from
The figure is from Frederico Azevedo, Suzana Herculano-Houzel and colleagues, published in 2009. They counted the cells in the brains of four adult men and got an average of 86.1 billion neurons (give or take 8.1 billion), plus about 85 billion other cells. Herculano-Houzel's review from the same year puts the average brain at about 1.5 kg.
Before that you'd mostly have read 100 billion neurons and ten times as many glial cells. Her review says those figures were usually quoted with no reference at all, and that when her group asked senior neuroscientists in 2007, none of them could name an original source. So I'd say 86 billion is better than what came before, even if it isn't the final word.
How you count them
Nobody counts 86 billion cells one by one. The method is called the isotropic fractionator, and as far as I understand it you turn a piece of brain into a soup of cell nuclei that's evenly mixed. You then count the nuclei in small samples, stain them to see which belong to neurons, and scale up to the whole brain. It's a density estimate, a bit like working out how many red blood cells you've got from a drop of blood and your total blood volume (that's the comparison the 2024 paper uses).
Where the neurons are
| Part of the brain | Mass | Neurons |
|---|---|---|
| Cerebral cortex | About 1,233 g | About 16 billion |
| Cerebellum | About 154 g | About 69 billion |
| Whole brain | About 1,508 g | About 86 billion |
What surprised me is that the cortex, the part you'd probably think of as doing the thinking, holds only about 19% of the neurons even though it's most of the weight. Most of them are in the cerebellum at the back, packed in as very small cells. The review also says the glia were never ten times the neurons. They're roughly equal in number.
How sure is anyone?
Not very, according to Alain Goriely, a mathematician at Oxford, writing in Brain in 2024. His point is mostly about sample size. If you take the four brains at face value, an average of 86.1 billion with a spread of 8.1 billion only lets you say the true average is somewhere between about 73 and 99 billion (95% confidence). He also got the data from a later study by Roberto Lent's group on five women aged 71 to 84, which gives about 61 to 73 billion. I re-did both intervals and got the same numbers.
So I wouldn't quote 86 billion as if it were exact. I'd say "somewhere between 61 and 99 billion, and about 86 billion in the best-known study", which is less catchy, I know. Goriely himself says most neuroscientists already treat 86 billion as an estimate.
What that means for wetware
I mostly care about this number because of the living-neuron computers I write about here, and I think it puts them in their place quite well. A Cortical Labs CL1 has about 800,000 human neurons on its chip, so if you take 86 billion at face value you'd need something like 107,500 of them to match one brain (and the brain organoids FinalSpark uses are smaller again). Your own brain does all of it on about 20 watts, which I've worked through on brain vs GPU power. If you'd like the history of the word, it's on what is wetware.
Questions
How many neurons are in the human brain?
You'll usually see about 86 billion, which comes from counting four men's brains in 2009. I'd take it as rough, though, since a 2024 paper reckons the data only back an average of somewhere between 61 and 99 billion.
Does the brain have 100 billion neurons?
Probably not. People used to quote 100 billion without saying where it came from, and when the 2009 team actually counted they got about 86 billion (they say 100 billion is outside the range they've found in human brains).
Are there more glial cells than neurons?
Not ten times more, though you'll still find that in some textbooks. The 2009 count found about 85 billion other cells to go with the 86 billion neurons, so it's closer to one each.
Sources
- Azevedo et al., Equal numbers of neuronal and nonneuronal cells, J. Comp. Neurol., 2009, as reported in Herculano-Houzel, The Human Brain in Numbers
- Eighty-six billion and counting: do we know the number of neurons in the human brain?
- Schottlender et al., Evaluating Encoding Strategies for Closed-Loop Classification in Biological Neural Networks, arXiv, 2026
- Balasubramanian, Brain power, PNAS, 2021