Guide · Autism
Is autism genetic? What causes autism, and what doesn't
Mostly, yes. Large family studies estimate that about 80% of the variation in autism comes from inherited genes, and it is not caused by parenting. Here is what the research shows, including on vaccines and paracetamol.
Updated 25 September 2026 · About 8 minutes to read · All guides
In short
- Autism is strongly genetic: twin and family studies of millions of children estimate that around 80% of the variation in autism is explained by inherited genes.
- There is no single autism gene. Most inherited likelihood comes from many common gene variants, each with a tiny effect, and genes come from both parents.
- When a family has one autistic child, about 1 in 5 younger siblings are also autistic, so about 4 in 5 are not.
- Older parental age and some pregnancy and birth complications are linked with a modest increase in likelihood. Parenting does not cause autism.
- Large studies have found no link between vaccines and autism, and no link with paracetamol in pregnancy when siblings are compared.
How much of autism is genetic?
Most of it. The largest studies used national health registers to compare how often autism occurs in identical twins, non-identical twins, siblings and half-siblings.
- Sweden: a study of children born between 1982 and 2006, including 37,570 twin pairs and more than 2.6 million full-sibling pairs, estimated the heritability of autism at 83%.
- Five countries, including Western Australia: a study of 2,001,631 children from Denmark, Finland, Sweden, Israel and Western Australia put heritability at about 81% (the middle of the country estimates, which ranged from 51% in Finland to 87% in Israel).
Heritability is easy to misread. It describes how much of the variation in autism across a whole population is explained by genetic differences. It does not mean an autistic parent has an 80% chance of having an autistic child.
Nor does it mean the rest is down to upbringing. In the Swedish study, the shared family environment (the home and upbringing siblings have in common) accounted for about 4%, with a range that included zero.
Is there an autism gene?
No. Many genes are involved and they act together. There are two broad kinds of genetic contribution.
- Common variants. Small DNA differences that are common in the general population, each shifting likelihood by a tiny amount. A large Swedish study found that most of the inherited likelihood of autism comes from this common variation.
- Rare changes, including new ones. Some are inherited. Others are de novo, meaning they appear for the first time in the child and are not carried by either parent. In the same Swedish study, new changes accounted for about 2.6% of the variation in autism likelihood, which is modest across the population.
Rare changes still matter for the people who have them. A 2020 study of 35,584 people identified 102 genes in which rare, disruptive changes raise the likelihood of autism. Changes in about half of those genes were found more often in people with significant developmental delay.
De novo changes are also one reason autism can appear in a family with no known history of it.
Will my next child be autistic?
Probably not, but the chance is higher than average. The international Baby Siblings Research Consortium followed 1,605 babies with an autistic older brother or sister, across 18 research sites, until age three. Overall, 20.2% were diagnosed autistic, similar to the 18.7% the group reported in 2011. About 4 in 5 were not.
| Group | Diagnosed autistic by age 3 |
|---|---|
| Younger sibling is a boy | 25.3% |
| Younger sibling is a girl | 13.1% |
| More than one older autistic sibling | 36.9% |
The chance was also higher when the older autistic child was a girl. The authors stress that these are group averages, not predictions for any one family. Your GP, paediatrician or a genetic counsellor can talk through your own family's situation.
Mother or father: who does autism come from?
There is no rule that autism comes from the mother or the father; genes come from both. The five-country study also tested for an extra ‘maternal effect’, beyond the genes a mother passes on, and found no support for one.
Parental age matters a little, on both sides. A study of 5.7 million children in five countries, including Western Australia, found that:
- children of fathers aged 50 or over were about 1.7 times as likely to be diagnosed autistic as children of fathers in their twenties
- children of teenage mothers were about 18% more likely, and children of mothers in their forties about 15% more likely, than children of mothers in their twenties
- likelihood rose as the age gap between the parents widened.
These are relative increases on a small base. In that study about 1 in 190 children were diagnosed autistic overall, and most children of older parents were not.
Pregnancy and birth: what is linked to autism
A 2017 review of the systematic reviews and meta-analyses on environmental factors found:
- Stronger links: older parental age, and birth complications involving trauma or a lack of oxygen.
- Weaker but real links: maternal obesity, maternal diabetes and caesarean birth.
- No link: vaccination, smoking in pregnancy and thimerosal (a preservative once common in vaccines). IVF and other assisted reproduction were most likely unrelated.
Raising Children Network also lists premature birth, poor growth in the womb and exposure to air pollution in pregnancy.
One medicine stands out. In a Danish study of 655,615 children, the 508 exposed in the womb to valproate, an epilepsy medicine, were diagnosed with autism at about three times the rate of the whole group (about 4.5% against 1.5%), even after allowing for the mother's epilepsy. Children whose mothers stopped valproate at least 30 days before conception did not have a raised risk. If you take valproate and are pregnant or planning a pregnancy, talk to your doctor before changing anything.
Most of these links are modest, and some may partly reflect genes that parent and child share rather than the exposure itself.
Do vaccines or paracetamol cause autism?
Vaccines
No. The evidence is large and consistent.
- A 2014 meta-analysis combined five cohort studies of 1,256,407 children and five case-control studies of 9,920 children. It found no link between vaccination and autism, including for the MMR vaccine, thimerosal and mercury.
- A 2019 Danish study followed 657,461 children born between 1999 and 2010, of whom 6,517 were diagnosed autistic. MMR-vaccinated children were not more likely to be autistic than unvaccinated children, and there was no increase in any subgroup, including children with an autistic sibling.
That last point matters if you are deciding about vaccines for a younger sibling. In 2025 the World Health Organization restated that childhood vaccines do not cause autism, and that studies claiming a link were flawed and have been discredited.
Paracetamol in pregnancy
In September 2025, claims made in the United States linked paracetamol in pregnancy (known there as acetaminophen) to autism. Australia's medicines regulator, the TGA, responded that paracetamol remains Pregnancy Category A in Australia, the category for medicines taken by large numbers of pregnant women without evidence of harm to the baby, and that it had no active safety investigation into paracetamol and autism.
The best evidence comes from comparing siblings. A Swedish study of 2,480,797 children found a very small association at first (about 5% higher likelihood of autism after exposure). When it compared brothers and sisters, where the mother used paracetamol in one pregnancy but not another, the association disappeared. A 2026 review of higher-quality studies, including sibling comparisons of 262,852 children assessed for autism, also found no increase, and put earlier concerns down to genetics or to the fever and pain paracetamol was taken for.
Genetic testing and autism assessment
Autism is diagnosed from a person's developmental history and how they communicate and behave, using standardised tools. It is not diagnosed with a genetic test.
Genetic testing is sometimes offered after or alongside a diagnosis to look for an underlying cause. It is arranged by a doctor, usually a paediatrician, sometimes with a clinical geneticist.
- Chromosomal microarray looks for missing or extra pieces of chromosomes. Medicare covers it for people with developmental delay, intellectual disability, autism, or two or more congenital abnormalities.
- Fragile X testing is commonly offered alongside it as a first-line test.
- Whole exome or genome sequencing is Medicare-funded for some children aged 10 or under, such as those with at least moderate intellectual disability or developmental delay, if a microarray found nothing.
For many families no specific cause is found. That fits what is known: much of the inherited likelihood comes from many small common variants rather than one identifiable change.
If your real question is whether your child is autistic, genetics won't answer it; a developmental assessment will. Our guide to signs of autism in children is a good place to start.
Questions people ask
Is autism hereditary or genetic?
Mostly both. ‘Genetic’ means genes are involved; ‘hereditary’ means passed down from parents. Twin and family studies estimate around 80% of the variation in autism is explained by inherited genes, which is why it runs in families. Some autistic people also have a new (de novo) genetic change that neither parent carries.
Who carries the autism gene, mother or father?
Neither specifically. There is no single autism gene, and the genes involved come from both parents. A study of two million children in five countries found no support for an extra ‘maternal effect’. Older age in either parent is linked with a modest increase in likelihood.
Is it my fault my child is autistic?
No. Healthdirect lists bad parenting, vaccinations, specific foods and paracetamol in pregnancy among things that do not cause autism. Most of the likelihood is inherited through many genes, and some changes appear for the first time in the child. The pregnancy and birth factors linked to autism are modest associations, not evidence that a parent caused it.
What causes autism during pregnancy?
No single pregnancy factor causes autism. Birth complications involving a lack of oxygen and older parental age have the clearest links, with weaker links for maternal obesity and diabetes. Valproate, an epilepsy medicine, taken in pregnancy is linked to a higher likelihood, though most exposed children are not autistic.
Do vaccines cause autism?
No. A meta-analysis covering more than 1.2 million children and a Danish study of 657,461 children found no link between vaccines, including MMR, and autism, including in children with an autistic sibling. The World Health Organization, Healthdirect and Raising Children Network all state that vaccines do not cause autism.
Does paracetamol cause autism?
The best evidence says no. A Swedish study of nearly 2.5 million children found no link when it compared siblings, and a 2026 review of higher-quality studies agreed. The TGA says paracetamol remains Pregnancy Category A in Australia. If you are pregnant, talk to your doctor or pharmacist before changing any medicine.
Wondering whether an assessment makes sense? The free screener takes about three minutes, for your child, your adult son or daughter, or yourself. It suggests which assessment fits, and it will tell you if one does not look like the right next step. Start the free screener.
Where this information comes from
We would rather you checked us than took our word for it. Every link below is the original source, not a summary of one.
- Sandin et al. (2017). The Heritability of Autism Spectrum Disorder. JAMA.Swedish population study of twins, siblings and half-siblings born 1982-2006 estimating autism heritability at 83%, with shared environment about 4%.
- Bai et al. (2019). Association of Genetic and Environmental Factors With Autism in a 5-Country Cohort. JAMA Psychiatry.Study of 2 million children in Denmark, Finland, Sweden, Israel and Western Australia estimating heritability at about 81% and finding no support for maternal effects.
- Gaugler et al. (2014). Most genetic risk for autism resides with common variation. Nature Genetics.Swedish genetic study finding most inherited autism likelihood comes from common variants, with new (de novo) mutations accounting for about 2.6% of liability.
- Satterstrom et al. (2020). Large-Scale Exome Sequencing Study Implicates Both Developmental and Functional Changes in the Neurobiology of Autism. Cell.Exome sequencing study of 35,584 people identifying 102 autism risk genes, about half more often affected in people with significant developmental delay.
- Ozonoff et al. (2024). Familial Recurrence of Autism: Updates From the Baby Siblings Research Consortium. Pediatrics.Prospective study of 1,605 infants with an autistic older sibling, finding 20.2% were autistic by age 3, with higher rates for boys and multiplex families.
- Sandin et al. (2016). Autism risk associated with parental age and with increasing difference in age between the parents. Molecular Psychiatry.Five-country study of 5.7 million children, including Western Australia, on how maternal age, paternal age and the parental age gap relate to autism.
- Modabbernia et al. (2017). Environmental risk factors for autism: an evidence-based review of systematic reviews and meta-analyses. Molecular Autism.Umbrella review grading the evidence for environmental factors, including parental age, birth complications, maternal obesity and diabetes, vaccination and smoking.
- Christensen et al. (2013). Prenatal valproate exposure and risk of autism spectrum disorders and childhood autism. JAMA.Danish population study of 655,615 children finding higher autism rates after exposure to valproate in pregnancy.
- Taylor et al. (2014). Vaccines are not associated with autism: an evidence-based meta-analysis of case-control and cohort studies. Vaccine.Meta-analysis of cohort studies of 1,256,407 children and case-control studies of 9,920 children finding no link between vaccination and autism.
- Hviid et al. (2019). Measles, Mumps, Rubella Vaccination and Autism: A Nationwide Cohort Study. Annals of Internal Medicine.Danish cohort of 657,461 children finding MMR vaccination did not increase autism risk, including in children with autistic siblings.
- Ahlqvist et al. (2024). Acetaminophen Use During Pregnancy and Children's Risk of Autism, ADHD, and Intellectual Disability. JAMA.Swedish study of 2,480,797 children in which a small association with paracetamol disappeared when siblings were compared.
- D'Antonio et al. (2026). Prenatal paracetamol exposure and child neurodevelopment: a systematic review and meta-analysis. The Lancet Obstetrics, Gynaecology, & Women's Health.Review of 43 studies prioritising sibling comparisons, finding no increased risk of autism, ADHD or intellectual disability after paracetamol in pregnancy.
- World Health Organization (2025). WHO statement on autism-related issues.WHO statement that there is no conclusive evidence linking paracetamol in pregnancy to autism and that childhood vaccines do not cause autism.
- ABC News (2025). Experts, TGA reject Trump statement autism is linked to paracetamol.Report quoting the TGA that paracetamol remains Pregnancy Category A in Australia, with comment from the AMA and RANZCOG on fever risks in pregnancy.
- Healthdirect (2025). Autism.Australian government health information on autism, including that it is linked to genes and environment and is not caused by vaccines, parenting, foods or paracetamol in pregnancy.
- Raising Children Network (2025). What causes autism?Australian parenting resource on genetic and environmental factors linked to autism and what does not cause it.
- Medicare Benefits Schedule. Item 73292: chromosome microarray.Medicare item for genome-wide microarray in people with developmental delay, intellectual disability, autism or at least two congenital abnormalities.
- Medicare Benefits Schedule. Item 73358: whole exome or genome sequencing.Medicare item for exome or genome sequencing in children aged 10 or under meeting set criteria, after a non-informative microarray.
- Centre for Genetics Education, NSW Health (2017). Fact Sheet 46: Autism Spectrum Disorder.Fact sheet noting paediatricians usually offer chromosomal microarray and fragile X testing, and that often no genetic cause is found.
Links to external sites are provided for reference. We do not control them and they may change. Last checked 25 September 2026.
This guide is general information, not medical advice. If you are worried about yourself or your child, talk to your GP.
Not sure where to start?
For your child, your adult son or daughter, or yourself. The free screener takes about three minutes and tells you which assessment fits — or whether you need one at all.