We know far less about the human brain than most people assume. Despite decades of neuroscience research, some of the most basic facts remain unsettled, including how many neurons the brain contains, whether it keeps producing new ones in adulthood, and what generates conscious experience. What we have learned, especially in the last 20 years, has repeatedly overturned older assumptions and revealed layers of complexity that weren’t even suspected a generation ago.
We Don’t Even Agree on the Neuron Count
For years, the standard figure was 100 billion neurons. That number was eventually challenged by a Brazilian neuroscientist who developed a new counting method and arrived at 86 billion, a figure that became the new consensus almost overnight. But a 2025 review published in the journal Brain examined the actual data behind these estimates and reached a blunt conclusion: we don’t know how many neurons the human brain contains. Experimental studies have returned wildly different results, ranging from 62 billion to 99 billion, with no overlap between some estimates and no clear explanation for the discrepancy. The variation between individual brains, differences in counting methods, and the sheer difficulty of working with postmortem tissue all contribute to the uncertainty.
If that sounds surprising for such a basic measurement, it illustrates a broader pattern in brain science. The brain is not easy to study while it’s alive, and the tools we have impose hard limits on what we can observe.
What We Can and Can’t See
The most widely used tool for studying the living brain is functional MRI, which tracks blood flow as a proxy for neural activity. It doesn’t measure neurons firing directly. Instead, it detects changes in blood oxygenation that follow neural activity with a delay of several seconds. That built-in lag means fMRI has poor temporal resolution. Without averaging multiple trials together, it can only distinguish between brain events separated by at least two to three seconds. Given that neurons fire on the scale of milliseconds, this is like trying to follow a conversation by checking in every few minutes.
Spatial resolution has improved over the years, but standard fMRI still groups together the activity of tens of thousands of neurons into a single data point. More advanced techniques can push finer, but there’s a fundamental trade-off: the closer you zoom in, the smaller the brain region you can study at once, and the more invasive the method tends to be. This is why so much of what we “know” about the brain comes in the form of which regions light up during a task, not what the individual circuits are actually computing.
The Cells We Overlooked for Decades
For most of modern neuroscience, glial cells were treated as passive scaffolding, the structural support that held neurons in place. That view has collapsed. Glia turn out to be active participants in brain signaling, waste removal, and immune defense, and they outnumber neurons in many brain regions.
Astrocytes, one major type of glial cell, produce calcium signals that fluctuate with wakefulness, peaking when you’re awake and dropping during sleep. These signals are most active in the fine branches of the cell, suggesting a level of local processing that researchers are still working to understand. Astrocytes also communicate with other glial cells through direct connections, passing calcium waves back and forth with oligodendrocytes (the cells that insulate nerve fibers). This means there’s an entire signaling network operating alongside the neuron-to-neuron communication that textbooks have traditionally focused on.
Microglia, the brain’s immune cells, have their own surprises. When the brain is under stress from neurodegenerative disease, microglia build temporary tunnels made of protein filaments and use them to transfer toxic protein clumps from overloaded cells to healthier neighbors. They even share mitochondria, the tiny power generators inside cells, through these tunnels. This kind of cooperative, on-demand networking was completely unknown until recently, and it suggests the brain’s immune system is far more organized than a collection of independent cells patrolling for damage.
The Storage Capacity Problem
Each neuron connects to thousands of others through synapses, and the total number of synaptic connections in the brain is estimated in the hundreds of trillions. A 2016 study from the Salk Institute discovered that synapses can vary in strength across at least 26 distinguishable sizes, not just the two or three categories previously assumed. This is the difference between a communication system that uses only zeroes and ones and one that uses an entire alphabet. Scaling that discovery up to the whole brain produced an estimate of roughly one petabyte of storage capacity, comparable to the entire World Wide Web at the time.
But storage capacity doesn’t tell you how the brain actually encodes, retrieves, or organizes information. We still don’t understand how a memory is physically stored, how the brain decides what to keep and what to discard during sleep, or why certain memories feel vivid decades later while others vanish within hours. The petabyte figure is useful as a benchmark, but it’s a measurement of potential, not a map of how the system works.
Does the Adult Brain Grow New Neurons?
One of the most contentious questions in neuroscience is whether the adult human brain continues to produce new neurons, a process called neurogenesis. In rodents and other mammals, the hippocampus (a region essential for memory) clearly generates new neurons throughout life. For years, the assumption was that the same applied to humans. Then, in 2018, two high-profile studies arrived at opposite conclusions using similar methods on human postmortem tissue. One found robust evidence of new neurons in people up to their 90s. The other found that neurogenesis drops to undetectable levels by adulthood.
The disagreement hasn’t been resolved. The difficulty lies in studying this process in human tissue after death, when the chemical markers used to identify young neurons degrade unpredictably depending on how the tissue was preserved and how long after death it was collected. This single question, whether your brain can grow new cells, remains genuinely unanswered.
The Gut-Brain Connection
One of the more surprising discoveries of recent decades is the extent to which the brain is influenced by the gut. The vagus nerve, the longest nerve in the body, runs from the brainstem to the abdomen and serves as a two-way communication highway. Stimulating vagal fibers in the gut directly affects brain systems involved in mood and anxiety. The gut’s own nervous system releases hormones like ghrelin that cross the blood-brain barrier and influence appetite, reward, and stress responses.
The trillions of bacteria living in your gut play a role too. In one study, people with PTSD had significantly reduced levels of three bacterial groups involved in immune regulation compared to trauma-exposed people who didn’t develop PTSD. Animal research has shown that exposure to certain immune-regulating bacteria can shift behavioral responses to stress toward more proactive coping. The mechanisms are still being mapped, but the implication is clear: the brain doesn’t operate in isolation. Its function is shaped by chemical signals originating far from the skull.
Consciousness Remains Unexplained
The deepest gap in our understanding is consciousness itself. We can identify brain regions associated with awareness, track changes in brain activity when someone loses and regains consciousness under anesthesia, and observe what happens during sleep stages. But we have no accepted explanation for why subjective experience exists at all, why there is “something it is like” to see the color red or feel pain.
At least five major theories compete for dominance. Global Neuronal Workspace Theory proposes that consciousness arises when information is broadcast widely across the brain, making it available to many systems at once. Integrated Information Theory takes a mathematical approach, arguing that consciousness corresponds to a specific measurable quantity related to how much a system integrates information beyond what its parts do individually. Higher-Order Theories suggest you’re only conscious of something when a separate brain process represents you as being aware of it. Recurrent Processing Theory focuses on feedback loops in sensory areas, and Predictive Processing frames consciousness as the brain’s ongoing effort to predict and correct its model of the world.
In 2022, proponents of all five theories participated in a public debate at the annual meeting of the Association for the Scientific Study of Consciousness. No consensus emerged. The theories make different predictions, and the experiments that could distinguish between them are only beginning to be designed. Consciousness is the clearest example of a phenomenon we experience every waking moment yet cannot explain in physical terms.
The Scale of What’s Unknown
Beyond the big questions, there are enormous categories of brain activity we’re only beginning to explore. Non-coding RNA, genetic material that doesn’t produce proteins, plays a role in brain development and function that likely represents only a small fraction of what the full RNA landscape is doing. The so-called “dark matter” of the genome is disproportionately active in the brain compared to other organs, and its functions are largely uncharacterized.
We also lack a complete wiring diagram. While projects like the Human Connectome Project have mapped large-scale fiber pathways between brain regions, a full neuron-by-neuron map of the human brain’s connections doesn’t exist. The only organism whose complete neural wiring diagram has been mapped is a roundworm with 302 neurons. The jump from 302 to something in the range of 70 to 90 billion, with hundreds of trillions of connections, is not just a matter of scale. It’s a fundamentally different kind of problem.
The honest summary is that we’ve learned enough to know the brain is vastly more complex than earlier models suggested. Nearly every “settled” fact, from the neuron count to the role of support cells to whether the brain regenerates, has been revised or challenged in the last two decades. What we know is impressive and growing. What we don’t know is larger.