Human brains are wrinkly because the cerebral cortex, the thin outer sheet of neural tissue, grows much faster than the deeper tissue it sits on, and this mismatch in growth rate forces the surface to buckle and fold. The process is surprisingly similar to what happens when you compress a thin, flexible sheet against a stiffer backing: it crumples. But while that comparison offers a quick mental image, the real story involves a choreography of cell division, mechanical stress, and developmental timing that researchers are still working to fully untangle.
A Fast-Growing Sheet on a Slow-Growing Core
The dominant explanation for cortical folding centers on differential growth. The cortex is a layered sheet of neurons only a few millimeters thick, sitting atop the white matter, which is the dense wiring that connects distant brain regions. During fetal development, the cortex expands in surface area far more rapidly than the white matter beneath it. Because the two layers are mechanically coupled, the faster-expanding cortex has nowhere to go but inward and outward, producing ridges (called gyri) and grooves (called sulci).
Computational models have shown that the resulting folding patterns depend on how much the cortex expands relative to its thickness and the overall brain size. A thinner cortex that expands a lot produces many tight folds; a thicker cortex that expands less stays relatively smooth. These relationships hold across a wide range of species, from nearly smooth mouse brains to the deeply convoluted brains of humans and dolphins.1PubMed Central. Gyrification from constrained cortical expansion The framework explains not just that brains fold, but why different brains fold to different degrees.
Mechanical models capture the two-component picture well: an expanding outer zone destined to become the cortex, coupled to an inner zone destined to become white matter that grows more slowly, perhaps partly in response to stress transmitted from the expanding surface above it.2PubMed Central. How Forces Fold the Cerebral Cortex The result is mechanical instability, and instability resolves itself as folds.
Timing and Biology Steer Where Folds Land
If folding were purely mechanical, you might expect folds to appear randomly, like wrinkles on a deflating balloon. But human brains fold in remarkably consistent patterns. The same major grooves show up in roughly the same locations in nearly every person. That consistency points to a biological layer on top of the physics.
Evidence supports a combined mechanism. Spatiotemporal patterns in how and when neurons are born and migrate bias where the primary folds form. Then the mechanical instability from cortical expansion takes over and propagates those primary folds, along with the smaller secondary and higher-order folds that give the brain its detailed texture.3PubMed Central. Mechanics of cortical folding: stress, growth and stability Think of it as biology choosing the general neighborhood of a fold and physics handling the fine-grained crumpling.
This interplay is visible during fetal development. The brain surface starts out smooth. Around the sixth month of gestation, the first primary grooves appear as short, isolated lines. As weeks pass, these grooves lengthen and branch, and secondary and tertiary grooves fill in between them. By birth, the brain has acquired its characteristic wrinkled appearance.4Acta Biomaterialia. Formation and evolution of gyri and sulci in fetal brains: A growth–release model
The Cells That Drive Expansion
The cortex doesn’t just passively stretch. It grows because progenitor cells in the developing brain divide furiously, producing the neurons that populate it. A specific type of progenitor cell called the outer radial glia, or oRG, has emerged as a central player. These cells are abundant in species with folded brains and rare in species with smooth ones, a pattern that strongly implicates them in cortical expansion and folding.5PubMed Central. A kinase-independent function of cyclin-dependent kinase 6 promotes outer radial glia expansion and neocortical folding
Research into primate-specific genes has reinforced this picture. When researchers introduced a gene called TMEM14B, which is active in primate oRG cells, into developing mouse brains, it triggered thickening of the cortex and the emergence of folds in an animal that would normally have a smooth brain. The gene boosted progenitor cell division, expanded the zone where oRG-like cells live, and led to proportional increases across all cortical layers.6PubMed. The Primate-Specific Gene TMEM14B Marks Outer Radial Glia Cells and Promotes Cortical Expansion and Folding The fact that a single gene could induce folding in a normally smooth brain underscores how tightly gyrification is coupled to the sheer number of neurons the cortex produces.
Why Pack So Much Surface Into a Small Space
The functional payoff of folding is straightforward: it allows a large cortical surface area to fit inside a skull that can still pass through a birth canal. The cortex is where the heavy lifting of perception, planning, language, and movement coordination happens, and the more cortical real estate you have, the more neural circuitry you can support. Folding is essentially a packing solution.
The folding pattern is not random decoration, either. In brain-mapping studies, researchers have found that the geometry of gyri and sulci reliably predicts the locations of distinct functional areas, especially in regions devoted to sensory processing. Primary and secondary sensory areas are extremely well predicted by the surrounding fold pattern, while the relationship becomes looser as you move into regions involved in more abstract cognitive tasks.7Cerebral Cortex. Cortical Folding Patterns and Predicting Cytoarchitecture Folding patterns, in other words, are not just cosmetic. They reflect the underlying architecture of neural circuits.
Do More Folds Mean More Intelligence?
This is one of the most commonly asked follow-up questions, and the answer is more nuanced than the pop-science version. Within humans, researchers have found that the degree of folding in certain cortical regions correlates with intelligence scores. In a study of 65 healthy adults, the strongest associations were in the left medial hemisphere, particularly a region in the temporo-occipital area near the posterior cingulate gyrus.8PubMed Central. Mapping the relationship between cortical convolution and intelligence: effects of gender
A larger study replicated the general pattern, finding that higher gyrification in a network spanning portions of the prefrontal cortex, the inferior parietal lobule, the temporoparietal junction, and several other regions was associated with greater general cognitive ability. The regional distribution was nearly identical across two independent samples, giving the finding more credibility than a one-off result.9PubMed Central. Regional Variations in Brain Gyrification Are Associated with General Cognitive Ability in Humans
But correlation is not destiny. The relationship is regional and statistical, not a simple “more wrinkles equals smarter.” A person with slightly less folding in one region is not doomed to lower cognitive performance. And across species, the link between gyrification and intelligence breaks down entirely in interesting ways. Some birds, particularly corvids and parrots, exhibit cognitive abilities comparable to those of great apes despite having non-cortical brains weighing as little as one to twenty-five grams, far smaller and structurally different from a primate brain.10PubMed Central. Why birds are smart Evolution has found more than one way to build a complex mind.
When Folding Goes Wrong
Because cortical folding depends on a chain of tightly coordinated events, disruptions at any stage can produce dramatically abnormal brains. Two conditions sit at opposite extremes.
Lissencephaly, which literally means “smooth brain,” occurs when neurons fail to migrate properly from their birthplace deep in the brain to their final positions in the cortex. Mutations in genes involved in the cytoskeleton, the internal scaffolding that cells use to move, are a common culprit. Without proper migration, the cortex cannot layer itself correctly, and the surface remains smooth or nearly so.11PubMed Central. Cytoskeleton in action: lissencephaly, a neuronal migration disorder Children with lissencephaly typically face severe developmental delays, seizures, and shortened life expectancy, illustrating how critical normal folding is to brain function.
At the other extreme is polymicrogyria, in which the cortex develops an abnormally large number of abnormally small folds. Rather than too few folds, the brain has too many, packed too tightly together. The condition appears to involve disruption of the brain’s outer surface during development, with defects in the protective membrane covering the brain, abnormal cell migration through those gaps, and changes to the blood supply in the surrounding tissue.12PubMed Central. Polymicrogyria: pathology, fetal origins and mechanisms Polymicrogyria can be localized to one region or widespread, and the severity of symptoms ranges accordingly, from mild motor difficulty to seizures and intellectual disability.
Folding Differences in Psychiatric and Developmental Conditions
Beyond the rare malformations, subtler differences in folding patterns have been detected in more common conditions. In autism spectrum disorder, brain imaging studies have found a mixed pattern of both increased and decreased gyrification compared to typically developing individuals, rather than a simple “more” or “less” story. A study drawing on over 300 individuals with autism and 350 controls across multiple sites found that these differences also followed unusual developmental trajectories, meaning the way gyrification changed with age differed between groups.13PubMed Central. Core-Symptom-Defined Cortical Gyrification Differences in Autism Spectrum Disorder
Schizophrenia, which also has strong neurodevelopmental roots, has similarly been linked to gyrification differences. Patients with schizophrenia and autism frequently experience motor abnormalities that may reflect atypical cortical development, and local gyrification measurements have been explored as a possible marker for these features.14PubMed. Local brain gyrification as a marker of neurological soft signs in schizophrenia These findings do not mean that folding patterns can diagnose anyone. But they suggest that the developmental processes driving gyrification overlap with the processes that, when disrupted, contribute to these conditions.
How Aging Unfolds the Brain
Cortical folding is not static after birth. As the brain ages, folds gradually flatten. Studies of older adults have found that gyrification decreases over the later decades of life, a pattern linked to the general atrophy that aging brains undergo. In a sample of 749 adults aged 55 to 85, age-related decreases in local gyrification were most prominent in the medial prefrontal cortex and the posterior cingulate cortex and precuneus, regions that form part of the brain’s default mode network.15PubMed. Age- and function-related regional changes in cortical folding of the default mode network in older adults
In Alzheimer’s disease, this age-related decline in folding accelerates. Research measuring gyrification alongside biomarkers of Alzheimer’s pathology has found that the disease-related pattern of decreased folding resembles an intensified version of what healthy aging does: cortical thickness drops, exposed surface area shrinks, and overall folding diminishes.16Scientific Reports. Cortical folding correlates to aging and Alzheimer’s Disease’s cognitive and CSF biomarkers Whether gyrification measurements could one day help detect Alzheimer’s disease earlier is an open question, but the relationship between folding loss and cognitive decline is real.
Does the Skull Squeeze the Brain Into Shape?
An older and intuitively appealing idea is that the skull constrains brain growth, forcing the cortex to fold the way a plant’s roots fold when they hit the walls of a pot. Computational models have lent some support to this: simulations show that a brain growing within a rigid boundary becomes more convoluted than one growing freely, suggesting the skull at least influences the folding process.17PubMed Central. A computational model of cerebral cortex folding
But experimental evidence has largely pushed back on the skull hypothesis. Classic experiments in which underlying brain tissue was removed to create space between the brain and skull showed that folding still occurred normally. And the skull itself is not a rigid cage during fetal development; it expands to accommodate brain growth rather than resisting it. For these reasons, most researchers have moved away from skull constraint as a primary driver of cortical folding.18Philosophical Transactions of the Royal Society B. Mechanics of cortical folding: stress, growth and stability The skull may fine-tune the outermost layer of convolutions, but the folding itself comes from within.
Measuring Folds and Mapping Them Across Populations
Quantifying how wrinkly a brain is turns out to be a surprisingly tricky measurement problem. The standard metric, the gyrification index, compares the total cortical surface area (including the buried portions hidden inside grooves) with the outer exposed surface. A perfectly smooth brain would have a gyrification index near 1.0, while a deeply folded human brain has a much higher value.
Researchers now measure this automatically from MRI scans. Software tools reconstruct the cortical surface from structural brain images with sub-millimeter accuracy, then generate an outer hull and compare the two surfaces across thousands of small regions, producing detailed maps of local gyrification for every patch of cortex.19PubMed Central. How to measure cortical folding from MR images: a step-by-step tutorial to compute local gyrification index Automated methods have been validated against manual measurement and shown to have equivalent accuracy, making large-scale comparisons across populations feasible.20PubMed. Automated computation of the Gyrification Index in prefrontal lobes: methods and comparison with manual implementation
These tools matter because they allow researchers to spot folding differences associated with conditions, aging, or cognitive traits that would be invisible to the naked eye on an MRI scan. Most of the clinical and cognitive findings described in the sections above rely on this kind of automated surface-based analysis.
Growing Folds in a Dish
One of the more striking recent developments has been the ability to grow miniature brain-like structures, called organoids, in the lab. These are tiny clumps of human neurons derived from stem cells that self-organize into layered structures loosely resembling early brain tissue. Under the right conditions, some organoids spontaneously develop surface folds, offering researchers a living model they can manipulate in ways that would be impossible in an intact brain.
Technical challenges remain. Organoids often suffer from poor oxygen delivery to their cores, leading to cell death at the center. Microfluidic devices that mimic aspects of the brain’s natural environment, including its extracellular matrix, have improved organoid survival and growth. In one approach, organoids grown in a microfluidic device with brain-derived matrix material showed roughly a quarter of the cell death seen in organoids grown in standard plate conditions, and reached volumes about four times larger.21Nature Communications. Microfluidic device with brain extracellular matrix promotes structural and functional maturation of human brain organoids Larger, healthier organoids are more likely to develop the kind of surface expansion that leads to folding, making them better tools for studying gyrification in a controlled setting. Whether these organoids can teach us something about human cortical folding that animal models cannot remains an active area of work, but the ability to watch folding happen in real time in a dish, and to genetically or physically perturb it, opens experimental doors that did not exist a decade ago.