86 Billion Neurons: The Real Story Behind the Number
The Number Everyone Gets Wrong
Quick: how many neurons does the human brain have?
If you said 100 billion, you just repeated one of the most persistent myths in all of neuroscience. It's in textbooks. It's in TED talks. It's on Wikipedia (well, it used to be). It gets cited by neuroscientists themselves, in academic papers, at conferences, in grant applications. One hundred billion neurons. A nice, round, satisfying number.
There's just one problem. Nobody actually counted. For decades, the "100 billion" figure was passed from textbook to textbook, professor to student, paper to paper, like a scientific game of telephone. And when someone finally did count, in 2009, the answer was different.
The real number is approximately 86 billion neurons. And the story of how we got from "100 billion (probably)" to "86 billion (actually counted)" involves dissolved brains, a stubborn Brazilian scientist, and a question so basic that it's embarrassing nobody thought to answer it sooner.
Why Didn't Anyone Just Count Them?
This seems like it should be straightforward, right? Take a brain, count the cells. We landed on the moon. We sequenced the genome. Surely we can count cells.
The problem is scale. A typical neuron's cell body is about 10-20 micrometers in diameter. The brain weighs about 1.4 kilograms. Counting 86 billion of anything is hard. If you counted one neuron per second, 24 hours a day, it would take you approximately 2,725 years.
For most of the 20th century, the standard approach was stereology: take a thin slice of brain tissue, count the neurons in that slice under a microscope, and then extrapolate to estimate the total. This method works reasonably well for small, relatively uniform brain regions. But the brain is not uniform. Neuron density varies wildly from region to region. The cerebral cortex has about 100,000 neurons per cubic millimeter in some areas. The cerebellum has roughly 2-3 million granule cells per cubic millimeter, making it one of the most densely packed structures in the body. Other regions are much sparser.
Extrapolating from small samples to the whole brain introduces enormous uncertainty. Different studies using different methods on different brain regions produced wildly different estimates. And because nobody was going to sit at a microscope for three millennia, the 100 billion figure stuck. It was roughly right. Roughly round. Roughly satisfying.
Then Suzana Herculano-Houzel decided she wanted an actual answer.
Dissolving Brains in Detergent: The Isotropic Fractionator
Herculano-Houzel, a neuroscientist at what was then the Federal University of Rio de Janeiro, had a problem. She wanted to compare neuron counts across different species, including humans. She couldn't do that with vague estimates. She needed a method that was fast, reliable, and could be applied to whole brains.
In 2005, she published a technique called the isotropic fractionator. The name is technical, but the method is beautifully simple in concept (and slightly horrifying in practice).
Here's what you do:
Step 1: Take the brain (or a specific brain region) and dissolve it. You mechanically dissociate the tissue and then immerse it in a detergent solution that breaks apart cell membranes. Everything dissolves except the cell nuclei, which are tough enough to survive. You're left with a suspension of free-floating nuclei in liquid, like cellular soup.
Step 2: Make the suspension uniform. Stir it until the nuclei are evenly distributed throughout the liquid. This is the "isotropic" part, meaning the distribution has no directional bias.
Step 3: Take a small, known fraction of the suspension and count the nuclei. You stain them with a fluorescent marker to make them visible, then count them under a microscope or with an automated counter.
Step 4: Multiply by the dilution factor. If you counted nuclei in 1/1000th of the total volume, multiply your count by 1,000 to get the total.
Step 5: Distinguish neurons from non-neurons. You use an antibody called NeuN that binds specifically to neuron nuclei. Stain your sample with NeuN, and you can count how many nuclei are neurons versus glia or other cell types.
The brilliance of this method is that it sidesteps the uniformity problem entirely. You don't need the brain to be uniform. You make the suspension uniform. Every nucleus is equally likely to be in any sample you take, regardless of where it originally sat in the brain. The statistical uncertainty drops dramatically.
Herculano-Houzel and her team applied this method to human brains. The result: 86 billion neurons. Not 100 billion. Fourteen billion neurons fewer than the textbooks claimed. That's a difference of about 14%, which in any other field of science would be considered a significant error.
Where Those 86 Billion Neurons Actually Live
The 86 billion is an average across adult human brains. But that total breaks down in ways that surprise most people.
| Brain Region | Neurons (approx.) | % of Total | Mass (% of brain) |
|---|---|---|---|
| Cerebral cortex | 16 billion | 19% | 82% |
| Cerebellum | 69 billion | 80% | 10% |
| Brainstem + other | ~1 billion | 1% | 8% |
Read that table again. The cerebellum, a structure the size of a small fist tucked under the back of the brain, contains 80% of all the neurons in your brain. The cerebral cortex, the wrinkled outer layer that takes up 82% of the brain's mass and gets virtually all the attention in popular neuroscience, has only 19%.
This was one of the most surprising findings to come out of Herculano-Houzel's work. For decades, neuroscientists had focused almost exclusively on the cortex as the seat of higher cognition. But the cerebellum has four times as many neurons.
Why? Because the cerebellum's neurons are tiny. Its granule cells are the smallest neurons in the brain, with cell bodies only about 5-8 micrometers across. They're packed at extraordinary density. The cortex's neurons, by contrast, are much larger (pyramidal neurons can have cell bodies of 20 micrometers or more) and are spread out with more space between them.
The cerebellum's 69 billion neurons are involved in motor coordination, balance, timing, and, as more recent research shows, cognitive functions like working memory and attention. But because its neurons are so small and its circuitry is relatively simple (compared to the cortex's complex, recurrent architecture), its massive neuron count doesn't translate to the same kind of computational flexibility.
EEG primarily detects the activity of the cerebral cortex, not the cerebellum. This is because cortical pyramidal neurons are large, aligned in parallel (perpendicular to the brain surface), and generate postsynaptic potentials that summate into detectable electrical fields. The cerebellum's granule cells are tiny, tightly packed, and oriented in ways that cause their electrical fields to largely cancel out. So while the cerebellum has 80% of the brain's neurons, it's the cortex's 16 billion that produce the brainwave patterns captured by EEG.
The "I Had No Idea" Moment: Your Brain Doesn't Scale Like Other Brains
Here's where Herculano-Houzel's work gets genuinely mind-bending.
For decades, neuroscientists assumed a simple scaling rule: bigger brains have more neurons. Makes intuitive sense, right? An elephant brain weighs about 5 kilograms, roughly three and a half times heavier than a human brain. So it should have proportionally more neurons.
It doesn't. Not even close.
Herculano-Houzel counted neurons in brains across dozens of species. What she found overturned a fundamental assumption. Brain size and neuron number don't have a consistent relationship across all mammals. They follow different scaling rules depending on which group of mammals you're looking at.
In rodents, as brains get bigger, neurons get bigger too and the space between them increases. A rat brain has about 200 million neurons. If you scaled a rodent brain up to human size using rodent scaling rules, it would contain only about 12 billion neurons and would weigh 35 kilograms. Obviously, that's not what happened.
In primates, something different evolved. As primate brains get bigger, neuron size stays relatively constant. Primate brains scale by adding more neurons without making each neuron dramatically larger. This means primate brains pack more computational power per gram than rodent brains.
The human brain exploits this primate scaling rule to its extreme. We have the most neurons of any primate, and our cerebral cortex has the most cortical neurons of any species studied to date. Not because our brains are the biggest (whales and elephants have bigger brains) but because our cortical neurons maintain a density that larger-brained non-primates can't match.
The Other Cells: Glia and the 1:1 Ratio
For every conversation about neurons, there's a parallel story about the brain's other cells: glia.
For most of the 20th century, the accepted wisdom was that the brain contained 10 glial cells for every neuron. This "10:1 ratio" was cited everywhere, often with a wink: "We have 100 billion neurons, supported by a trillion glia." It was a nice narrative. Neurons were the stars. Glia were the support crew, outnumbering the stars 10 to 1.
Herculano-Houzel's method destroyed this myth too.
Using the isotropic fractionator, her team counted both neuron and non-neuron nuclei in human brains. The ratio of glia to neurons turned out to be approximately 1:1, not 10:1. The human brain has roughly 85 billion non-neuronal cells (mostly glia) alongside its 86 billion neurons.
Where did the 10:1 myth come from? Nobody knows for sure. It may have been an extrapolation from brain regions where glia do vastly outnumber neurons (this is true in some subcortical areas). Or it may have been a number someone estimated once, and like the 100 billion neurons figure, it just got repeated until it felt like fact.
The corrected ratio has implications. Glia, it turns out, aren't just passive support cells. Astrocytes actively modulate synaptic transmission. Oligodendrocytes produce the myelin that speeds up neural communication. Microglia serve as the brain's immune system. The 1:1 ratio suggests the brain invests as many cells in its support infrastructure as in its computational machinery. That's not a staffing ratio you'd expect if glia were mere janitors.
Why Does the Number Matter?
You might wonder why it matters whether the brain has 86 billion neurons or 100 billion. Fourteen billion neurons is a lot, sure, but in the context of tens of billions, does the precision change anything?
It changes how we understand what makes the human brain special.
If you accept the old numbers uncritically (100 billion neurons, 10:1 glia ratio, bigger brains always mean smarter animals), you end up with a story that doesn't actually explain human cognition. Elephants and whales have bigger brains. If brain size were the whole story, they should be smarter than us. They're not. Or at least, not in the ways we typically measure intelligence.
Herculano-Houzel's accurate counts tell a different story. What makes the human brain exceptional isn't raw size. It's cortical neuron number. We have more neurons in our cerebral cortex than any other species, roughly 16 billion. This is because primate brains evolved a more efficient scaling rule that lets them pack more neurons into less space. And the cortex is where the brain does its most flexible, abstract, creative thinking.
The accurate count also reframes comparisons between humans and other primates. A chimpanzee's brain has about 6 billion cortical neurons. Ours has 16 billion. That's roughly a 2.7-fold difference. Is a 2.7-fold increase in cortical neurons enough to explain language, mathematics, art, science, and everything else that separates human cognition from chimp cognition? Maybe. But it at least gives us the right numbers to work with.
What 86 Billion Neurons Look Like to EEG
You can't see 86 billion neurons individually with any current technology (except, in principle, electron microscopy of fixed tissue, one painstaking slice at a time). But you can see what they do collectively.
When millions of cortical neurons fire their postsynaptic potentials in synchrony, the summed electrical field is strong enough to pass through the skull and reach electrodes on the scalp. This is EEG. And the patterns it captures, the delta, theta, alpha, beta, and gamma oscillations, reflect the coordinated behavior of vast neural populations.
The Neurosity Crown sits at 8 positions across the scalp (CP3, C3, F5, PO3, PO4, F6, C4, CP4), each one listening to the synchronized chorus of millions of cortical neurons in the region beneath it. At 256Hz, it takes 256 snapshots of that chorus per second. The N3 chipset handles the signal processing on the device itself, turning raw electrical signatures into frequency band power, focus scores, calm scores, and data streams accessible through JavaScript and Python SDKs.
You'll never count the neurons in your own brain. But you can listen to them. You can watch their collective behavior shift in real-time as you focus, relax, meditate, or think. And that changes your relationship with the number entirely. 86 billion neurons isn't just a trivia answer. It's the machinery behind every experience you've ever had.
The Numbers in Perspective
Let's end with some scale, because the numbers deserve it.
86 billion neurons. Each one forms roughly 7,000 synaptic connections. That's approximately 600 trillion synapses. If each synapse were a star, your brain would contain more stars than six Milky Way galaxies.
Those 86 billion neurons consume about 20% of your body's total energy, despite the brain representing only about 2% of your body weight. Per gram, the brain is roughly 10 times more metabolically active than the rest of the body.
Each neuron can fire anywhere from once per second to several hundred times per second. At any given moment, billions of neurons are firing in complex, interlocking patterns that encode everything you perceive, think, feel, and remember.
For most of human history, all of this was invisible. Locked inside bone. Inaccessible except in death.
Now you can put a device on your head and watch 16 billion cortical neurons do their thing. Not individually, but collectively, as the waves and rhythms that constitute your conscious experience. That's not 100 billion, and it's not a guess. It's 86 billion, counted by a scientist who dissolved brains in detergent because she wanted the real answer.
Science doesn't get much better than that.