Autism is strongly genetic. Twin studies consistently estimate that 64% to 91% of autism risk comes from inherited factors, making it one of the most heritable neurodevelopmental conditions. But “genetic” doesn’t mean simple. There isn’t a single autism gene. Instead, hundreds of genes contribute in different combinations, and environmental factors during pregnancy can shape how those genes are expressed.
How Much of Autism Risk Is Genetic?
The most reliable way to measure genetic influence is by studying identical twins, who share 100% of their DNA, and comparing them to fraternal twins, who share about 50%. A large meta-analysis of twin studies found that genetic factors account for 64% to 91% of the likelihood of developing autism. The remaining risk comes from shared environmental influences, which account for roughly 7% to 35% depending on how narrowly autism is defined.
That wide range reflects something important: autism isn’t a single condition with a single cause. It’s a spectrum with significant variation in how it presents, what genes are involved, and how much environment plays a role in any given person.
Common Variants vs. Rare Mutations
Genetic risk for autism comes in two broad flavors, and they work very differently.
The first is common genetic variants, meaning small differences in DNA that are widespread in the general population. Individually, each one has a tiny effect. But collectively, common variants may explain up to half of autism’s heritability. Think of it like height: no single gene makes you tall, but thousands of small genetic nudges add up. Most autistic people likely carry a unique combination of these common variants, which is one reason the condition looks so different from person to person.
The second type is rare mutations, often called de novo mutations because they arise spontaneously in a child rather than being passed down from a parent. These can have a large effect on an individual level, sometimes disrupting a single gene enough to substantially increase autism risk on their own. Despite their power in individual cases, all de novo mutations combined explain only about 2% of the overall variation in autism risk across the population. They’re high-impact but uncommon.
Over 100 Genes Are Linked to Autism
Researchers have identified at least 126 genes that pose a substantial risk for autism. These genes tend to be involved in how brain cells connect and communicate with each other, particularly during early development. Some affect how synapses (the junctions between neurons) form, while others influence the proteins that help brain cells stick together and pass signals.
In roughly 20% to 25% of autistic children and adults, a specific genetic cause can be identified. This includes de novo mutations, rare inherited variants, and chromosomal abnormalities. For the remaining 75% to 80%, no single genetic culprit can be pinpointed. Their autism likely results from a complex mix of many common variants, each contributing a small amount of risk.
Syndromic Forms of Autism
A subset of autism cases are linked to known genetic syndromes. Fragile X syndrome, Rett syndrome, and MECP2 duplication syndrome are among the most recognized. These are caused by mutations in a single gene or chromosomal region, and they produce a broader pattern of symptoms beyond autism traits alone. Children with these syndromes often have intellectual disability, seizures, or distinctive physical features alongside their autistic characteristics. These syndromic forms are relatively rare but are among the most genetically well-understood causes of autism.
Siblings Have a 1 in 5 Chance
Family studies offer some of the most practical evidence for autism’s genetic basis. Data from the Baby Siblings Research Consortium shows that about 20.2% of younger siblings of an autistic child will also develop autism. That’s roughly ten times the rate in the general population.
Interestingly, the sex of the older autistic sibling matters. Younger siblings of autistic girls are more likely to develop autism than younger siblings of autistic boys. This fits a pattern researchers call the “female protective effect,” where girls appear to need a higher genetic load to develop autism. When a girl does develop autism, it suggests she carries more risk variants, which means her siblings are more likely to inherit enough of those variants to cross the threshold themselves.
Environmental Factors Still Matter
Genetics doesn’t tell the whole story. Several prenatal factors are associated with increased autism risk: advanced age in either parent, gestational diabetes, bleeding during pregnancy, and being a firstborn child. These aren’t causes in the way a gene mutation is a cause. They’re factors that appear to interact with existing genetic susceptibility.
One of the clearest examples of how environment and genes interact involves valproate, an anti-seizure medication. When a fetus is exposed to valproate in the womb, the drug changes how genes are read and expressed, a process called epigenetics. Specifically, it alters chemical tags on DNA and the proteins that package it, which can turn genes on or off at the wrong times during brain development. Some of the genes affected by valproate exposure overlap with genes already linked to autism, including those involved in synapse formation and the connections between brain cells.
Animal research has shown this works in the other direction too. In mice genetically predisposed to autism-like behaviors, an enriched social environment with more space and more interaction with other mice reduced the severity of those behaviors. This suggests that even when a strong genetic risk is present, environment can dial the effects up or down.
Can a Genetic Test Predict Autism?
Not yet, and possibly not ever as a standalone tool. Genetic testing is useful after a diagnosis to look for a specific underlying cause, especially in children who have autism alongside intellectual disability or other medical features. Current guidelines recommend chromosomal microarray testing and screening for Fragile X syndrome as a starting point. Whole exome sequencing, which reads the protein-coding portions of all genes at once, is increasingly used and tends to identify a genetic cause more often than older methods.
What genetic testing cannot do right now is predict whether a person will develop autism before symptoms appear. Researchers have developed polygenic risk scores, which add up the effect of many common gene variants into a single number. In theory, a higher score would mean higher risk. In practice, these scores are not ready for clinical use. A systematic review found that while polygenic scores show a statistical association with autism diagnosis at a population level, they are too inconsistent and imprecise to be meaningful for any individual person. The scores also have a major equity gap: nearly all the research behind them is based on people of European ancestry, making them even less reliable for other populations.
Autism is also influenced by rare genetic variants, epigenetic changes, and environmental factors that polygenic scores don’t capture. Even with future improvements, a genetic score alone is unlikely to ever provide a complete picture of individual risk. Any future clinical application would need to be part of a much broader assessment.
What This Means for Families
If you’re a parent of an autistic child wondering whether genetics played a role, the answer is almost certainly yes. Genes are the single largest contributor to autism risk. But the genetics are complex: hundreds of genes are involved, they interact with each other and with prenatal environment, and the specific combination differs from family to family. Having one autistic child does raise the probability for future children to about 1 in 5, but it also means 4 out of 5 younger siblings will not be autistic.
For autistic adults curious about their own genetics, clinical testing can sometimes identify a specific genetic variant, particularly if autism co-occurs with other medical or developmental features. For many people, though, the genetic architecture behind their autism is a mosaic of common variants that current technology can’t fully map. The science is advancing quickly, but the complexity of autism’s genetic landscape means simple answers remain rare.