IQ is substantially genetic, but it is far from entirely so. The best estimates from twin and family studies put the heritability of intelligence at roughly 50 to 80 percent in adults, meaning that somewhere between half and four-fifths of the variation in cognitive ability across a population traces back to inherited DNA differences. That leaves a large role for environment, and the interplay between genes and surroundings turns out to be more interesting than either factor alone. How much your genes matter for your IQ depends, paradoxically, on what kind of environment you grew up in.
What Twin Studies Actually Tell Us
Most of what we know about the heritability of intelligence comes from comparing identical twins (who share virtually all their DNA) with fraternal twins (who share about half). When identical twins score more alike on IQ tests than fraternal twins do, the gap is attributed to genetics. The landmark Minnesota Study of Twins Reared Apart took this further by tracking identical twins who had been separated in infancy and raised in different homes. Even under those conditions, about 70 percent of the variation in IQ was associated with genetic differences.1PubMed. Sources of human psychological differences: the Minnesota Study of Twins Reared Apart That finding made a strong case that genes carry real weight, because the twins did not share a household, neighborhood, or school system.
But twin studies come with caveats. Twins reared apart are not placed into random homes; adoption agencies historically matched families on background characteristics, which can inflate the apparent genetic contribution. And “heritability” itself is a population-level statistic. It describes how much of the spread in scores across a group is linked to genetic variation in that group, under the environmental conditions that group actually experienced. Change the environment and you change the heritability estimate. This is not a technicality; it reshapes the whole conversation.
Why Heritability Rises with Age
One of the more counterintuitive findings in intelligence research is the Wilson Effect: the heritability of IQ climbs as people get older. In early childhood, shared family environment accounts for a sizable chunk of the differences between children. By the late teens and into adulthood, heritability reaches roughly 0.80 and stays there.2PubMed. The Wilson Effect: the increase in heritability of IQ with age
The reason is not that your genes suddenly become more powerful at age eighteen. It is that as people gain more control over their own lives, they tend to gravitate toward environments that match their genetic predispositions. A child with a natural inclination toward reading ends up in book clubs, advanced classes, and eventually a college major that reinforces that inclination. A child without that inclination drifts toward other activities. Over time, the environments people select amplify the genetic differences that were always present, while the shared family environment they had no choice about (the same house, the same parents, the same schools) recedes in influence.
Thousands of Genes, Each Doing Almost Nothing
If intelligence were controlled by a handful of genes, finding them would have been straightforward. Instead, intelligence turns out to be what geneticists call “highly polygenic,” meaning thousands of DNA variants each contribute a tiny nudge.3PubMed Central. Polygenic Scores for Cognitive Abilities and Their Association with Different Aspects of General Intelligence—A Deep Phenotyping Approach Recent large-scale genome studies have identified inherited sequence differences accounting for about 20 percent of the roughly 50 percent heritability of intelligence.4Europe PMC. The new genetics of intelligence That means the majority of the genetic contribution is still unaccounted for at the molecular level, scattered across the genome in variants too small to detect individually.
Researchers now aggregate these tiny effects into what are called polygenic scores, summing up thousands of genetic nudges into a single number that predicts cognitive performance to a modest degree. These scores also show overlap with physical health outcomes: modest but widespread genetic correlations exist between intelligence and various health conditions.5PubMed Central. What genome-wide association studies reveal about the association between intelligence and physical health, illness, and mortality That overlap hints that some of the same biological pathways, perhaps involving neural efficiency or cellular maintenance, influence both cognitive performance and general health. But the prediction power of polygenic scores remains limited. They explain a small fraction of the variation in any one person’s IQ, and they are far from being a crystal ball.
How Poverty and Wealth Change the Equation
Here is where the nature-versus-nurture framing starts to break down. A landmark study of young children in the United States found that in impoverished families, roughly 60 percent of the variation in IQ was explained by the shared home environment, and the contribution of genes was close to zero. In affluent families, the result flipped almost exactly: genes dominated and shared environment contributed little.6Psychological Science. Socioeconomic Status Modifies Heritability of IQ in Young Children The same children, born with the same DNA, would show different heritability estimates depending on the household they landed in.
This pattern, sometimes called the Scarr-Rowe hypothesis, holds that genetic potential for intelligence can only express itself fully when basic environmental needs are met. If nutrition is poor, if the home is chaotic, if lead paint is peeling off the walls, those environmental harms swamp whatever genetic advantages a child might carry. Remove those barriers and genetic differences become the main source of variation. A study of German twins found support for this in middle childhood: family socioeconomic status amplified genetic effects during those years, with shared environment explaining more of the variation at the lower end of the income spectrum.7PubMed Central. Socioeconomic status amplifies genetic effects in middle childhood in a large German twin sample
But this interaction is not universal. A cross-national analysis found clear gene-by-socioeconomic-status effects in U.S. studies, but in Western European countries and Australia, where social policies provide more uniform access to education and healthcare, the interaction was essentially zero or even reversed.8PubMed Central. Large Cross-National Differences in Gene × Socioeconomic Status Interaction on Intelligence In other words, the degree to which poverty suppresses genetic potential depends on the social safety net. In countries where the floor is higher, family income matters less for cognitive development, and genes account for more of the variation even among poorer families. The gene-environment interaction is itself shaped by policy.
Environmental Factors That Measurably Shift IQ
A number of environmental exposures have well-documented effects on cognitive ability, independent of genetics. Lead is one of the clearest examples. A meta-analysis found that an increase in blood lead from 10 to 20 micrograms per deciliter was associated with a drop of about 2.6 IQ points, and the researchers concluded these associations should be considered causal.9PubMed. Low-level lead exposure and children’s IQ: a meta-analysis and search for a threshold Lead disrupts signaling between brain cells, particularly in the hippocampus, a region critical for learning and memory. For children in older housing with deteriorating lead paint, or communities with contaminated water, this is not a subtle statistical effect. It is a concrete, preventable drag on cognitive development.
Nutrition, iodine deficiency, prenatal alcohol exposure, and chronic stress during early childhood all have similar documented impacts. These are not rare exotic hazards; they are common features of disadvantaged environments worldwide. The practical implication is straightforward: reducing environmental harms is one of the most reliable ways to raise cognitive outcomes at the population level, because these harms disproportionately affect children who are already at a disadvantage.
The Flynn Effect and What It Proves
Perhaps the strongest evidence that environment matters comes from the Flynn Effect, the well-documented rise in average IQ scores across generations throughout the twentieth century.10PubMed Central. The Flynn effect: a meta-analysis A formal meta-analysis covering over a century of data (1909 to 2013), with nearly four million participants from 31 countries, confirmed substantial worldwide gains.11PubMed. One Century of Global IQ Gains: A Formal Meta-Analysis of the Flynn Effect (1909-2013) Genetic change cannot explain gains this rapid, because the human gene pool does not shift meaningfully over a few decades. Improvements in nutrition, education, sanitation, and the increasing cognitive demands of modern life are the leading explanations.
Intriguingly, several countries have seen the Flynn Effect slow down or reverse in recent decades. A Norwegian study using military conscription data spanning birth cohorts from 1962 to 1991 found that both the original rise and its subsequent decline could be fully accounted for by within-family variation, meaning that siblings born in different years within the same family showed the same pattern.12PubMed Central. Flynn effect and its reversal are both environmentally caused That analysis controlled for everything shared between siblings and found no evidence that genetic factors or between-family environmental differences were driving the trend. Whatever is causing scores to rise and then plateau appears to be something that changes over time within the same families, such as shifts in educational practices, media exposure, or other cultural factors.
How the Brain Reflects These Influences
Brain imaging studies have started to reveal the neural architecture underlying intelligence. One finding that holds up across studies is that more intelligent individuals tend to have brain networks organized in a more efficient pattern, resembling what network scientists call a “small-world” structure, where information can travel between distant brain regions through relatively few steps. Research has identified the parietal cortex as a major hub in this network, with greater connectivity in that region associated with higher intelligence scores.13PubMed Central. Functional brain network efficiency predicts intelligence
This kind of neural efficiency is itself influenced by both genes and experience. Brain development during childhood and adolescence responds to stimulation, nutrition, and stress in ways that shape how efficiently networks form. Cortical thickness, the depth of the brain’s outer layer, also correlates with intelligence and changes dynamically during development.14PubMed Central. The Dynamic Associations Between Cortical Thickness and General Intelligence are Genetically Mediated So even at the level of brain structure, genes and environment are intertwined. Your DNA sets parameters for how the brain develops, but experience determines where within those parameters you end up.
Can You Train Your Way to a Higher IQ?
The commercial brain-training industry would love the answer to be yes. But the evidence is disappointing. A study that had healthy young adults practice working memory tasks intensively found that participants improved substantially on the specific tasks they trained on, and those gains lasted at least six months. However, there was no transfer to untrained tasks, fluid intelligence, or other cognitive abilities.15PLOS ONE. Failure of Working Memory Training to Enhance Cognition or Intelligence You get better at the game you practice, but that improvement does not spill over into general smarts.
This does not mean IQ is fixed for life. Education, particularly sustained high-quality education during childhood, has measurable effects on cognitive test scores. The problem is that many interventions show promising initial gains that fade once the program ends. Researchers have identified that lasting impacts require either continued skill-building, opening doors to new opportunities (a “foot-in-the-door” effect), or placement in environments that sustain the gains. A one-off brain game cannot replicate the conditions under which real cognitive development occurs.
Why Group Heritability Claims Don’t Hold Up
One of the most persistent misunderstandings in this field is the assumption that because IQ is heritable within a population, differences in average scores between populations must also be genetic. This reasoning is formally incorrect. A recent paper demonstrated that knowing the heritability of a trait within groups provides no information about the heritability of differences between groups.16PubMed Central. Heritability within groups is uninformative about differences among groups: Cases from behavioral, evolutionary, and statistical genetics Even if between-group heritability could be estimated, it would not reveal whether genes or environment are responsible for the gap, or even in which direction each factor pushes.
The classic illustration involves plants grown in two different soils. Within each plot, differences in height are entirely genetic (same soil, different seeds). But the difference in average height between the two plots is entirely environmental (different soils, same seed mix). Knowing that height is 100 percent heritable within each plot tells you nothing about whether the between-plot difference is genetic. The same logic applies to human populations that have experienced very different environmental conditions, including different histories of nutrition, education, discrimination, and environmental exposures like lead.
When “Culture-Free” Tests Are Not Free of Culture
Even the measurement side of this question is complicated. IQ tests designed to be culturally neutral, using abstract visual puzzles rather than language, still produce skewed results when norms from one country are applied to children in another. A study testing Moroccan children with a nonverbal “culture-free” intelligence test found dramatic misclassification when British norms were used: up to about 16 percent of healthy children scored in the “intellectually impaired” range, and up to 62.5 percent fell “below average.”17Archives of Clinical Neuropsychology. Cultural Bias in Intelligence Assessment Using a Culture-Free Test in Moroccan Children These were healthy children with no cognitive impairment, misclassified because the scoring benchmarks came from a different cultural context.
The misclassification rates worsened at older ages, suggesting that the cultural gap in test performance widens as children progress through different educational and social systems. This finding underscores that IQ scores are not a pure readout of biological cognitive capacity. They reflect a mix of innate ability, educational exposure, test familiarity, and cultural context. Any discussion of whether IQ is “genetic” has to grapple with the reality that the thing being measured is itself shaped by environment in ways that are hard to fully separate out.
Genetic Overlap with Health and Psychiatric Conditions
The genes associated with intelligence do not operate in isolation. Genome-wide studies have revealed that intelligence shares genetic architecture with a surprisingly wide range of outcomes, including educational attainment, mental health conditions, and physical health. The same common genetic variants that nudge cognitive ability slightly upward or downward also correlate with risk for conditions like schizophrenia and with measures like body mass index and blood pressure.18Egyptian Journal of Medical Human Genetics. Cataloging the potential SNPs (single nucleotide polymorphisms) associated with quantitative traits, viz. BMI (body mass index), IQ (intelligence quotient) and BP (blood pressure): an updated review
This genetic overlap does not mean that being smart protects you from heart disease, or that high blood pressure makes you less intelligent. It means the biological systems underlying these traits, brain development, vascular health, metabolic function, share some of the same genetic plumbing. Polygenic scores developed for educational attainment, for instance, also predict small portions of the variance in verbal memory and crystallized intelligence among people with major psychiatric disorders.19PubMed Central. The genetic relationship between educational attainment and cognitive performance in major psychiatric disorders The picture emerging from molecular genetics is one of deeply interconnected biology, where “intelligence genes” are really just genes that happen to affect brain function among many other things.
This interconnectedness also helps explain why the “missing heritability” gap has been so stubborn. Twin studies suggest heritability of around 50 to 80 percent, but molecular methods can currently account for only a fraction of that. Part of the gap comes from rare variants that large studies are not powered to detect, part from gene-gene interactions that are difficult to model, and part from the sheer number of tiny effects spread across the genome. Comparing molecular results to twin study results provides insights into the genetic architecture of intelligence, but the full map is still being drawn.20Molecular Psychiatry. Genetics and intelligence differences: five special findings