Is intelligence fixed at birth? Explore what twin, adoption, schooling and nutrition research really shows about genes, environment, the Flynn effect, and what can and cannot change.
Last updated: 12 October 2026. Written by the EDUSAMBAM Editorial Team. This article is for general education and does not replace advice from a qualified professional.
Many people believe that your IQ is decided at birth and stays the same for life. Others believe that with enough hard work anyone can become a genius. The scientific evidence supports neither extreme. Studies of twins, adopted children, school reforms, nutrition programmes and whole populations across a century show that genes matter a great deal, that environment matters too, and that the two work together. In this article you will learn what the best research actually found, in plain language, with the sources named so that you can check them yourself.
IQ scores are fairly stable from person to person across life, but they are not frozen. Genes explain a substantial part of the differences between people, and that share is larger in adults than in children. Environment still matters: average scores rose by about three points per decade through much of the twentieth century, an extra year of schooling raises scores by roughly one to five points, and severe nutritional or social deprivation can lower them by a lot. Early education programmes have shown lasting benefits for life outcomes even when IQ gains fade. No quick trick or app can reliably raise intelligence, but good health, good schooling and learning habits can help.
"Can IQ change?" is really three different questions, and mixing them up causes most of the confusion.
| Question | What it asks | Short answer |
|---|---|---|
| Stability | If I rank high or low compared with others of my age now, will I rank similarly later? | Mostly yes, especially from adolescence on, but with real individual variation |
| Causes | How much of the difference between people comes from genes and how much from environment? | Both contribute; genes explain a larger share as people get older |
| Changeability | Can a person, or a whole population, raise scores through changed conditions? | Yes, to a degree, as shown by schooling, nutrition and the Flynn effect |
Notice that a trait can be both highly inherited and changeable by environment at the same time. Human height is a good example. Height differences between people in the same country are strongly genetic, yet average height rose sharply in many countries over the last century as nutrition and health improved. The same logic applies to cognitive test scores.
Before asking whether IQ changes, we must ask how well it is measured. Full-length individually given tests such as the Wechsler scales are very reliable, which means that a person tested twice in a short period gets similar scores. But no measurement is perfect, and a score always has a margin of error.
| Source of change | What happens | What it means |
|---|---|---|
| Measurement error | A person's score on one day may be a few points above or below their "true" score; psychologists report a confidence range, often about 4 to 6 points either side | Small differences between two scores are usually not meaningful |
| Practice effects | Repeating the same test within weeks usually raises scores by a few points because the tasks are familiar | A higher second score does not always mean a higher ability |
| Different tests | Two tests sample different tasks and use different norm groups | Scores from different tests are not exactly interchangeable |
| Day-to-day state | Tiredness, anxiety, illness, hunger and low motivation can lower performance | A score reflects performance on one occasion |
| Real change | Learning, health, development and ageing change true ability | This is the change researchers want to understand |
So when you hear that someone's IQ "changed by ten points", ask whether it was real change or the normal wobble of measurement. Both happen, and the best studies try to separate them.
To separate the effects of genes and environment, scientists use natural experiments. The two most important are twin studies and adoption studies.
Large reviews of twin studies have found that identical twins' IQ scores are strongly correlated, at about 0.86 on average, while fraternal twins' scores are correlated at about 0.60. A well-known study of identical twins who were separated early and raised apart (the Minnesota Study of Twins Reared Apart, 1990) found a correlation of about 0.70, which is remarkably high given their different homes.
| Pair | Share genes? | Share home? | Typical IQ correlation |
|---|---|---|---|
| Identical twins raised together | Almost all | Yes | About 0.86 |
| Identical twins raised apart | Almost all | No | About 0.70 |
| Fraternal twins raised together | About half of varying genes | Yes | About 0.60 |
(A correlation of 1.0 means a perfect match and 0 means no relationship.) These results are one of the reasons why psychologists accept that genes play a major role in differences in cognitive ability.
There is no single "intelligence gene". Large genetic studies, including a study of over a million people in 2018, have found that thousands of tiny genetic differences each have a very small effect. When these are added together into a "polygenic score", the score explains only a modest share of differences in cognitive test performance (roughly 7 to 10 per cent in that study), so it cannot predict what any one person will achieve.
Heritability is a statistic that estimates what proportion of the differences between people in a particular group, in a particular environment, is linked to genetic differences. It is easily misunderstood, so keep these four points in mind.
Imagine a handful of seeds that differ genetically, some from tall plants and some from short plants. If you plant them all in the same rich soil, any height differences you see will be almost entirely genetic. But if you plant one seed in rich soil and another from the same bag in poor soil, the second grows shorter, though its genes are the same. Genes and environment are both always at work.
One of the most surprising findings is that the heritability of cognitive ability increases with age. A study of about 11,000 pairs of twins from four countries (Haworth and colleagues, 2010) estimated it at 41% at age 9, 55% at age 12 and 66% at age 17. Later research suggests that it may be even higher in adulthood.
Why would this happen? One leading idea is that as children grow older, they have more freedom to choose and shape their own experiences: a child with a natural love of puzzles seeks out more puzzles, which strengthens that ability. Younger children's experiences are chosen more by parents and teachers. This does not mean that environment stops mattering, only that people increasingly select and create environments that suit their genetic tendencies.
A well-known 2003 study of American twins (Turkheimer and colleagues) reported that in very poor families, the environment seemed to explain more of the differences in children's IQ and genes explained little, while in better-off families, the opposite was found. This suggested that deprivation can hold children back from reaching their genetic potential. Later studies have found this pattern more clearly in the United States than in some other countries, so scientists treat it as an important idea that still needs more research.
If environment did not matter, then changing it would not change scores. Researchers have tested this in many different ways. The table below summarises some of the best-known findings.
| Environmental factor | What research found | Source |
|---|---|---|
| Schooling | Each additional year of education raised IQ scores by about 1 to 5 points (average near 3 to 4) in a combined analysis of 42 datasets with over 600,000 people; the effect persisted into later life | Ritchie and Tucker-Drob, 2018 |
| Iodine in the diet | In studies from China, children in severely iodine-deficient areas scored about 12 points lower on average; different reviews estimate a gap between roughly 7 and 13 points. Iodine before and during pregnancy recovered about 9 points | Qian and colleagues, 2005 |
| Institutional care | Children raised in orphanages averaged an IQ of 84, compared with 104 for children raised in families, a 20-point gap, in a review of 75 studies from 19 countries | van IJzendoorn, Luijk and Juffer, 2008 |
| Adoption | Adopted children scored substantially higher than siblings or peers who stayed behind in poor conditions, and similar to their new environmental peers | van IJzendoorn, Juffer and Klein Poelhuis, 2005 |
| Early education | Better school and life outcomes in randomised studies such as the Abecedarian Project (see Section 8) | Campbell and colleagues, 2012 |
These are average results from groups, and individual outcomes vary widely. But the message is consistent: a healthy body, enough nutrition, a safe and stimulating home, and good teaching all help the brain develop to its potential. In extreme neglect or severe malnutrition, the effects can be large.
Salt iodisation is one of the cheapest public health measures in the world. Adding a tiny amount of iodine to table salt helps prevent the thyroid and brain-development problems linked to iodine deficiency, and it is a major reason why governments promote iodised salt.
In the 1980s the New Zealand political scientist James Flynn noticed something puzzling. When IQ tests are re-standardised so that the average is again 100, the older versions turn out to be "easier" for today's people. In other words, people in many countries were scoring higher on the same questions than their parents and grandparents had. A famous 1987 paper reported gains in 14 nations.
The largest analysis so far (Pietschnig and Voracek, 2015) combined 271 samples, covering almost 4 million people from 31 countries between 1909 and 2013. It found gains of roughly 3 points per decade overall. The gains were different for different kinds of test, as the table shows.
| Type of test | Estimated gain per decade |
|---|---|
| Fluid intelligence (new problems, patterns) | About 4 points |
| Spatial ability | About 3 points |
| Full-scale IQ | About 3 points |
| Crystallised intelligence (knowledge, vocabulary) | About 2 points |
The same analysis found that gains were weaker in recent decades than earlier in the century.
A gain of 30 points over a century would suggest that an average person in 1920 would score about 70 today, which sounds absurd. The authors of the large analysis themselves argued that it is unlikely to mean this, and that the gains may partly reflect changes in how people approach the tests. Flynn himself suggested that modern life, with more schooling and more abstract and symbolic thinking, trained people to be better at the kind of reasoning that tests ask for, such as spotting similarities and classifying ideas, even if it did not make them wiser.
Because genes cannot change across a few generations, the Flynn effect must have environmental causes. Researchers have proposed several, and probably more than one is true.
| Proposed cause | Reasoning |
|---|---|
| Better nutrition and health | Healthier childhoods, fewer infections and better prenatal care help brain development |
| More and better schooling | People now spend more years in education, which is known to raise test performance |
| More abstract thinking in daily life | Modern work and media use classification, symbols and hypothetical reasoning |
| Smaller families and more attention per child | Parents may be able to give each child more time and resources |
| Greater test familiarity | People are more used to tests, puzzles and standard question styles |
Evidence suggests that the gains were strongest on tests of fluid reasoning and weaker on knowledge tests, which fits the idea that changes in thinking habits and health, not just more facts, played a part.
In some wealthy countries, the rise has slowed, stopped or even reversed. The clearest evidence comes from Norway. A 2018 study by Bratsberg and Rogeberg analysed the conscription test results of over 700,000 Norwegian men born between 1962 and 1991. Scores rose for those born until about 1975 and then declined for those born later. The key finding was that the rise and fall could be seen within families, when comparing brothers, which rules out explanations based on changes in who had children. The authors concluded that both the rise and the decline were caused by environmental factors. What exactly changed in the environment is still debated, and the pattern is not the same in every country.
The Flynn effect and its reversal are strong evidence that average test scores respond to the world around us. They show that population averages are not fixed, even when individual differences within a generation are partly genetic.
If early environment matters, then high-quality early education should help. Two famous American experiments tested this.
| Programme | What was done | Findings |
|---|---|---|
| Abecedarian Project (North Carolina, started in the 1970s) | Of 111 infants from low-income families, children were randomly assigned to a full-time, year-round, high-quality educational childcare programme from infancy until they entered kindergarten, or to a comparison group | At age 30, the programme group had more years of schooling (about 13.5 compared with about 12.3) and were about four times as likely to hold a four-year college degree (23% compared with 6%) |
| Perry Preschool Project (Michigan, 1960s) | Low-income children aged 3 to 4 randomly assigned to a quality preschool programme with home visits, or to no programme | Test-score gains faded within a few years, but the programme group did better later in school completion, employment and avoiding trouble with the law, according to follow-ups to age 40 |
Notice an important pattern: early gains in test scores often fade after the programme ends, yet the children still do better in life. Researchers think this may be because the programmes build motivation, learning habits, social skills and schooling success, not only raw test performance. A fair note of caution: these programmes were small, intensive and expensive, so results from large public programmes can be smaller.
Scores from the first years of life are poor predictors of later scores. They become more stable through childhood and the teenage years. Yet change is still possible. A 2011 study in the journal Nature (Ramsden and colleagues) tested 33 healthy teenagers at about 14 and again about four years later. Their verbal IQ scores changed by between about minus 20 and plus 23 points, and changes in the scores matched changes in the density of grey matter in particular brain regions. This was a small study, and some of the change may be measurement error, but it supports the idea that the teenage brain is still developing.
Rank-order stability is high over very long periods. In the Lothian Birth Cohort 1921, the same test taken at ages 11 and 90 was correlated at 0.54 (0.67 after correcting for the fact that only the healthiest and brightest survive to be tested at 90). Other follow-ups from the Scottish Mental Surveys found that about half of the differences between people were stable between age 11 and old age. This means that, on average, a child who ranked high among classmates is likely to rank high among peers in old age, but that rank can still change substantially for individuals.
At the same time, average raw performance on speeded, new-problem tasks (fluid intelligence) tends to decline in later adulthood, while knowledge and vocabulary (crystallised intelligence) hold up much longer, as we explained in our article on Types of Intelligence.
| Life stage | Stability of rank among peers | Typical changes |
|---|---|---|
| Infancy and early childhood | Low to moderate | Rapid development; tests are weak predictors |
| School age | Moderate to high | Schooling raises scores; genetic influence grows |
| Teenage years | High, but individual changes happen | Verbal and non-verbal abilities can shift |
| Adulthood | Very high | Knowledge keeps growing; fluid skills slowly decline |
| Older age | High over decades | Slower processing; health, activity and education matter |
Here is an honest summary of the evidence.
| Approach | What research suggests |
|---|---|
| More education and challenging study | The most consistent effect found; roughly 1 to 5 points per extra year in studies of schooling, and benefits that appear to last |
| Practising the same test | Raises scores a little, but this is familiarity with the test, not a real gain in ability |
| Commercial "brain-training" games | People improve at the games they practise. A major review of the evidence (Simons and colleagues, 2016) found little good evidence that this carries over to wider thinking skills |
| Good sleep, nutrition and physical activity | Support healthy brain function and learning; they are good for the brain even if they do not add large numbers of IQ points |
| Learning knowledge and skills deeply | Builds crystallised intelligence and makes new problems easier to solve |
| Quick tricks, music for babies, special diets or supplements for healthy people | No reliable evidence of large, lasting IQ gains |
The practical lesson is that the most reliable route is the oldest one: keep learning, keep studying and look after your health. A few IQ points are less important than gaining knowledge, skills and good habits that make you capable.
The psychologist Carol Dweck popularised the idea of a growth mindset, the belief that abilities can be developed through effort and good strategies, as opposed to a fixed mindset, the belief that abilities are set. It is appealing, and believing you can improve certainly helps you to keep trying.
It is worth being accurate about the evidence, however. A large 2018 meta-analysis (Sisk and colleagues) found that, overall, the links between mindset and school achievement, and the effects of short mindset interventions, were weak. The interventions seemed to help more for students at risk of doing poorly and for students from lower-income backgrounds. So a growth mindset is a useful habit but it is not a magic switch. What helps most is a growth mindset combined with good teaching, good study methods and real opportunities to practise.
Because IQ is such a sensitive topic, scientists urge caution on several points.
| Myth | Fact |
|---|---|
| IQ is completely fixed at birth | Scores are fairly stable in rank, but schooling, health and nutrition can change them |
| IQ is mostly environment, so anyone can reach any level | Genes explain a substantial part of individual differences, and effort cannot erase them |
| "60% heritable" means 60% of your intelligence is genetic | It describes differences within a group, not the make-up of one person |
| There is an "intelligence gene" | Thousands of genetic variants each contribute a tiny amount |
| The Flynn effect proves people today are far smarter than their grandparents | It shows that test performance has changed, probably through health, schooling and thinking habits, but it does not mean a gain in overall wisdom |
| Early education programmes always raise IQ permanently | Test-score gains often fade, though life-outcome benefits have been found in some studies |
| Brain-training apps make you generally smarter | Improvement on the trained tasks is common, but wider transfer is not well supported |
| A growth mindset alone raises grades | Effects are small on average and work best alongside good teaching and strategies |
| If you are... | Helpful actions backed by the evidence |
|---|---|
| A student | Attend school regularly; use retrieval practice and spaced study; read widely; sleep well; ask for help early |
| A parent or carer | Talk, read and play with children from infancy; provide healthy food; protect sleep; praise effort and strategies, not just results |
| A teacher | Teach knowledge explicitly; give frequent low-stakes quizzes; give feedback; avoid labelling children by a score |
| An adult learner | Keep learning new, challenging things; stay physically active; stay socially connected; manage health conditions such as hearing loss |
| A community or policy maker | Support nutrition, health care, quality schools and early childhood education |
Finally, remember the key idea of this article: IQ is a useful but limited measurement of one kind of ability, and it responds to the conditions in which people live and learn. This is a hopeful message, because conditions can be improved.
The main claims in this article come from the following published studies and reviews. You can search for them by author and year.
| Topic | Source |
|---|---|
| Heritability increases with age | Haworth, C. M. A., and colleagues (2010). The heritability of general cognitive ability increases linearly from childhood to young adulthood. Molecular Psychiatry, 15, 1112-1120. |
| Twins reared apart | Bouchard, T. J., and colleagues (1990). Sources of human psychological differences: The Minnesota Study of Twins Reared Apart. Science, 250, 223-228. |
| Poverty and heritability | Turkheimer, E., and colleagues (2003). Socioeconomic status modifies heritability of IQ in young children. Psychological Science, 14, 623-628. |
| Genetic studies of cognition | Lee, J. J., and colleagues (2018). Gene discovery and polygenic prediction from a genome-wide association study of educational attainment in 1.1 million individuals. Nature Genetics, 50, 1112-1121. |
| Education raises IQ | Ritchie, S. J., and Tucker-Drob, E. M. (2018). How much does education improve intelligence? A meta-analysis. Psychological Science, 29, 1358-1369. |
| Iodine | Qian, M., and colleagues (2005). The effects of iodine on intelligence in children: a meta-analysis of studies conducted in China. Asia Pacific Journal of Clinical Nutrition, 14, 32-42. |
| Orphanage care | van IJzendoorn, M. H., Luijk, M. P. C. M., and Juffer, F. (2008). IQ of children growing up in children's homes: A meta-analysis on IQ delays in orphanages. Merrill-Palmer Quarterly, 54, 341-366. |
| Adoption | van IJzendoorn, M. H., Juffer, F., and Klein Poelhuis, C. W. (2005). Adoption and cognitive development. Psychological Bulletin, 131, 301-316. |
| The Flynn effect | Flynn, J. R. (1987). Massive IQ gains in 14 nations. Psychological Bulletin, 101, 171-191. Pietschnig, J., and Voracek, M. (2015). One century of global IQ gains. Perspectives on Psychological Science, 10, 282-306. |
| Reverse Flynn effect | Bratsberg, B., and Rogeberg, O. (2018). Flynn effect and its reversal are both environmentally caused. PNAS, 115, 6674-6678. |
| Early education | Campbell, F. A., and colleagues (2012). Adult outcomes as a function of an early childhood educational program: An Abecedarian Project follow-up. Developmental Psychology, 48, 1033-1043. Schweinhart, L. J., and colleagues (2005). Lifetime Effects: The HighScope Perry Preschool Study Through Age 40. |
| Teenage IQ change | Ramsden, S., and colleagues (2011). Verbal and non-verbal intelligence changes in the teenage brain. Nature, 479, 113-116. |
| Stability to old age | Deary, I. J., Pattie, A., and Starr, J. M. (2013). The stability of intelligence from age 11 to age 90 years: The Lothian Birth Cohort of 1921. Psychological Science, 24, 2361-2368. Deary, I. J., and colleagues (2000). Intelligence, 28, 49-55. |
| Brain training | Simons, D. J., and colleagues (2016). Do "brain-training" programs work? Psychological Science in the Public Interest, 17, 103-186. |
| Growth mindset | Sisk, V. F., and colleagues (2018). To what extent and under which circumstances are growth mind-sets important to academic achievement? Psychological Science, 29, 549-571. |
| Term | Meaning |
|---|---|
| Correlation | A number from 0 to 1 showing how closely two sets of scores go together |
| Flynn effect | The rise in average IQ test scores across generations during much of the twentieth century |
| Heritability | The share of differences between people in a group that is linked to genetic differences, in a given environment |
| Identical twins | Twins who share essentially all their genes |
| Meta-analysis | A study that combines results from many earlier studies |
| Polygenic score | A number that adds up the small effects of many genetic variants |
| Practice effect | A rise in score because a person has seen the test or similar tasks before |
| Randomised study | A study in which participants are assigned to groups by chance, which allows fair comparison |
| Reliability | How consistently a test gives similar results when repeated |
Intelligence test scores are shaped by the lives we live as well as by the genes we carry. Rank among peers is fairly stable, yet whole populations have risen and fallen, schooling adds points, and a good start in life can change what people go on to achieve. Be sceptical of anyone who says that intelligence is fixed and nothing can be done, and equally of anyone who promises a quick fix. Healthy bodies, good schools, curious minds and steady practice are still the best investment.
14 questions. Select an answer for each, then submit to see your score instantly.