Intelligence is partly genetic. Twin and adoption studies estimate that genetic differences account for about half or more of the variation in IQ among adults in the populations studied, with most estimates between 50 and 80 percent, and for less in childhood. Environment accounts for the rest. Heritability describes differences within a population, not how much of one person's IQ "comes from" genes, and a highly heritable trait can still be changed by the environment.
The details matter, because "genetic" is often heard as "fixed" or "explains group differences." Neither follows from the research.
What heritability actually means
Heritability is a statistic: the share of the variation in a trait, within a particular population at a particular time, that is associated with genetic differences among its members. A heritability of 50 percent for IQ does not mean half of your IQ points come from your genes and half from your upbringing. Genes and environments don't add up that way inside a single person.
Three consequences are often missed:
- Heritable is not unchangeable. Height is highly heritable, yet average height rose substantially over the 20th century as nutrition and health improved. IQ scores also rose across generations (the Flynn effect), far too fast for genetic change.
- Heritability depends on the environment. If everyone grew up in identical conditions, all remaining differences would be genetic and heritability would approach 100 percent. Where environments vary widely, environment explains more.
- Heritability within groups says nothing about differences between groups. A trait can be highly heritable within each of two populations while the gap between them is entirely environmental. The mainstream scientific position is that heritability estimates cannot be used to explain average differences between countries or ethnic groups; the factors researchers can actually demonstrate behind such differences are schooling, health, nutrition, economic conditions and familiarity with tests.
How twin and adoption studies work
Identical twins share essentially all their DNA; fraternal twins share about half of the genetic variants that differ between people, like ordinary siblings. If identical twins are more alike in IQ than fraternal twins raised in the same home, the extra similarity points to genetic influence. Comparing the two kinds of twins lets researchers split variation into three parts: genetic, shared environment (what makes siblings in the same home alike, such as family income or parenting style), and non-shared environment (experiences unique to each person, plus measurement error).
Adoption studies add a different lens. Adopted children can be compared with both their biological and adoptive families. Studies of identical twins raised apart, most famously the Minnesota Study of Twins Reared Apart led by Thomas Bouchard, found such twins strikingly similar in IQ despite growing up in different homes.
Adoption studies also show the environment at work. In a Swedish national study, Kendler and colleagues (2015) compared brothers in which one was raised by the biological parents and the other was adopted into a different home. The adopted brothers, who on average went to better-educated families, scored about 4 IQ points higher than their brothers raised at home. Both findings hold at once: genes matter, and the family environment can shift scores.
Twin studies rest on assumptions, notably that identical and fraternal twins experience equally similar environments. Critics have questioned those assumptions, but the broad conclusion that IQ is substantially heritable has held up across many designs and countries.
Why heritability rises with age
One of the most consistent findings in the field is that the heritability of IQ increases from childhood into adulthood. This is sometimes called the Wilson effect, after Ronald Wilson's work on the Louisville Twin Study (Bouchard, 2013). Plomin and Deary (2015) describe heritability rising from about 20 percent in infancy to perhaps 80 percent in later adulthood.
| Age | Approximate heritability of IQ | Source |
|---|---|---|
| Infancy | About 20% | Plomin & Deary (2015) |
| Childhood (around 9) | About 41% | Haworth et al. (2010) |
| Early adolescence (around 12) | About 55% | Haworth et al. (2010) |
| Late adolescence (around 17) | About 66% | Haworth et al. (2010) |
| Later adulthood | Up to about 80% | Plomin & Deary (2015) |
The Haworth figures come from about 11,000 twin pairs in four countries. Over the same period, the influence of shared family environment shrinks, and by adulthood it is small in most studies.
The leading explanation is gene–environment correlation. Children are first placed in environments by their parents. As they grow up, they increasingly choose and shape their own: what they read, which friends and activities they seek out, how much school they pursue. Those choices partly reflect genetic dispositions, so small initial differences get amplified into larger ones. In this picture, genes often work through environments rather than around them.
What DNA studies show
There is no single "IQ gene." Genome-wide association studies, which scan millions of genetic variants in hundreds of thousands of people, have found thousands of variants linked to intelligence and educational attainment, each with a tiny effect. Adding them up gives a polygenic score.
Current polygenic scores capture a modest share of differences in intelligence, about 4 percent of the variance in the review by Plomin and von Stumm (2018), and somewhat more in later studies, well below the twin-study estimates. That shortfall is known as the "missing heritability." Two further caveats matter:
- Scores shrink within families. Comparing siblings, polygenic scores predict noticeably less than they do across unrelated people. Part of the population-level prediction reflects family environment correlated with parents' genes, sometimes called genetic nurture, rather than a direct effect of a child's own DNA.
- Scores are not portable. Most genetic studies have used samples of European ancestry, and the scores predict less accurately in other populations for technical reasons such as differences in how genetic markers are correlated. This is a limitation of the tool, not evidence about group differences.
Polygenic scores can't tell you a person's IQ. They are useful as research tools for studying development in large samples, not for predicting or selecting individuals.
Does family income change heritability?
In an influential study of 7-year-old twins in the United States, Turkheimer and colleagues (2003) found that heritability depended on socioeconomic status. In the poorest families, most of the variation in IQ was explained by shared environment and heritability was close to zero. In affluent families, the pattern reversed and genetic differences explained most of the variation. The interpretation: when environments are deprived, they hold children back and swamp genetic differences; when environments are supportive, genetic differences have more room to show.
That finding has not replicated everywhere. A meta-analysis by Tucker-Drob and Bates (2016) found the gene-by-SES interaction in US samples but not in samples from Western Europe and Australia, where the effect was near zero or slightly reversed. One proposed reason is that stronger social safety nets and more equal access to schooling and health care narrow the range of deprivation children experience. That explanation is plausible but not proven. The result is a reminder that heritability is a property of a population in its circumstances.
How much does environment matter?
A heritability of 50 to 80 percent in adulthood still leaves substantial room for environment, and some environmental effects are well documented:
- Schooling. A meta-analysis by Ritchie and Tucker-Drob (2018) estimated that each additional year of education raises IQ by about 1 to 5 points. See IQ and education.
- Family environment, as the Swedish adoption study above shows, particularly when it differs sharply in resources.
- Health and nutrition, which are among the leading candidate explanations for rising scores across generations.
Much of the "environment" estimate in twin studies is non-shared: experiences unique to each person, including chance events, illness and specific opportunities. These are hard to pin down, but they are real. For what this means for change over a lifetime, see whether IQ can change over time.
What this means for you
Your genes are one input, not a ceiling. IQ is fairly stable in adulthood, and much of that stability reflects genes and environments working together. But heritability statistics describe populations, not your individual potential, and they don't tell anyone what you can learn or achieve. For parents, the practical takeaway is unglamorous: a stable, stimulating home, good schooling, and attention to health give children's abilities the best conditions to develop, whatever their genetic starting point. If you are concerned about a child's development, a pediatrician or school psychologist is the right first stop; this article is not medical advice.
For more on what IQ does and doesn't measure, see what IQ is.