WHO estimates that about 1 in 127 people worldwide had autism in 2021.
Autism spectrum disorder is a neurodevelopmental condition characterized by differences and difficulties in social communication and interaction, alongside restricted or repetitive patterns of behavior, interests or activities. Abilities and support needs vary.
Autistic people may be more sensitive or less sensitive to sensory input, such as sound, light or temperature; sensory experiences vary between individuals.
In DSM-5, autism spectrum disorder includes several previously separate diagnoses, including Asperger’s disorder.
A 2020 experimental study by Crompton and colleagues found more effective information transfer in same-neurotype than mixed-neurotype groups of adults. However, a larger 2025 replication found no such difference. These task-specific findings do not establish a universal rule about autistic communication.
Keep reading to learn more about the prevalence and statistics of children with autism.
The following figures describe different populations and reporting years; they should not be treated as directly comparable estimates.
- WHO estimates that about 1 in 127 people worldwide had autism in 2021.
- CDC surveillance identified autism in about 1 in 31 eight-year-olds at 16 U.S. sites in 2022.
- CDC’s 2018 surveillance identified autism spectrum disorder (ASD) in about 1 in 44 eight-year-olds at 11 U.S. sites.
- CDC’s ADDM estimates for eight-year-olds were 6.7 per 1,000 in 2000 and 18.5 per 1,000 in 2016; participating sites and methods varied.
- In CDC’s 2022 ADDM surveillance, identified autism prevalence among eight-year-olds was 3.4 times as high in boys as in girls.
- CDC’s monitored communities recorded about 1 in 150 eight-year-olds with autism in 2000 and 1 in 31 in 2022; sites and methods differed.
- Recorded prevalence varies by surveillance area and assessment methods; these data do not establish a ranking of all U.S. states.
- Planning the transition from school to adult health, education and employment services is important; support needs do not end when school finishes.
- Employment estimates depend on the population, reporting period and definition of employment; unemployment and underemployment are different outcomes.
- Some autistic children use few or no spoken words. Communication needs should be assessed individually, including appropriate augmentative and alternative communication.
- Autistic children can have co-occurring conditions, including ADHD, anxiety and depression; estimates depend on the population and assessment method.
- Motor difficulties can co-occur with autism. Questionnaire screening for motor impairment is not equivalent to a confirmed diagnosis.
- In a U.S. study published in 2012, 55.1 percent of autistic young adults had held paid employment during the first six years after high school. This was an observed outcome, not a prediction of lifetime employment.
- Leigh and Du’s 2015 modeling study projected annual U.S. medical, nonmedical and productivity costs of $268 billion for 2015 and $461 billion for 2025. These were historical projections, not measured spending totals or current estimates.
- A 2014 study by Buescher and colleagues estimated annual U.S. societal costs of $175–196 billion for adults and $61–66 billion for children, using 2011 prices. These modeled totals included services and productivity losses and depended on assumptions about intellectual disability.
- Healthcare-cost estimates vary by age, co-occurring conditions, insurance coverage and study methods; a single spending multiplier should not be applied to all autistic children.
Recorded autism prevalence has increased in CDC surveillance. Changes in identification, diagnostic practices, records and participating sites affect interpretation; these figures do not by themselves establish a change in underlying occurrence.
Autism Prevalence Numbers
- WHO’s global all-age estimate for 2021 was about 1 in 127
- Available estimates do not support a reliable worldwide ranking of countries.
- Low recorded prevalence may reflect limited identification or incomplete data.
- U.S. surveillance-site estimates should not be presented as a ranking of entire states.
Autism Prevalence According to the CDC
The CDC monitors autism in the United States. They have released information on the evolving research on autism diagnosis and examined many study papers.
CDC surveillance helps identify differences in autism identification and access to assessment. Its findings highlight continuing needs for equitable diagnostic and support services; they do not establish that access to care is already equal.
CDC’s ADDM estimates for eight-year-olds were 6.7 per 1,000 in 2000 and 18.5 per 1,000 in 2016; participating sites and methods varied
CDC’s 2022 ADDM estimate was about 1 in 31 eight-year-olds at 16 U.S. sites; this measures identified prevalence, not lifetime probability.
CDC’s 2022 ADDM surveillance identified autism 3.4 times as often among eight-year-old boys as girls.
In a U.S. parent-report study covering 2009–2017, about 17 percent of children aged 3–17 had a diagnosed developmental disability, including autism, ADHD, blindness or cerebral palsy.
According to the CDC, autism occurs in all socioeconomic, racial, and ethnic groups.
This is a ratio of identified prevalence in a specified surveillance population, not an individual prediction.
Extensive research shows that childhood vaccines do not cause autism.
Autism Rates By Year
For surveillance year 2000, CDC’s ADDM Network estimated autism prevalence at 6.7 per 1,000 eight-year-olds, or about 1 in 150, across six participating sites.
For 2004, ADDM estimated autism prevalence at 8.0 per 1,000 eight-year-olds, or about 1 in 125, across eight sites.
ADDM estimates were about 1 in 110 eight-year-olds at 11 sites in 2006 and 1 in 88 at 14 sites in 2008.
For 2010, ADDM estimated autism prevalence at about 1 in 68 eight-year-olds across 11 sites. Participating areas and methods varied across surveillance years.
In 2018, CDC surveillance identified autism in approximately 1 in 44 eight-year-olds, or 23 per 1,000, at 11 U.S. sites. This is about 3.4 times the 2000 ADDM estimate; sites and methods differed.
Awareness, diagnostic practices and access to assessment can influence how many autistic people are identified, including people whose autism was previously missed.
Diagnostic definitions and identification practices have changed over time, limiting direct comparisons between historical prevalence studies.
Greater awareness and access to assessment can improve identification. However, many autistic people still experience unmet healthcare, educational and support needs.
Childhood vaccines do not cause autism. The paper that promoted a link between the MMR vaccine and autism was fraudulent and was retracted; its author subsequently lost his medical license.
Population growth can increase the number of autistic people without increasing prevalence, which is the proportion of a defined population identified as autistic. Estimates should always specify the population, methods and reporting year.
Early autism studies used diagnostic definitions and case-finding methods that differ from those used today.
Changes in diagnostic definitions are one consideration when interpreting historical increases in recorded prevalence.
In 2018, CDC’s ADDM Network identified autism in 230 per 10,000 eight-year-olds across 11 U.S. sites.
A precise historical fold-change should not be inferred from studies with different populations, definitions and methods.
Autism Prevalence By Sex Among Eight-Year-Olds at 11 U.S. ADDM Sites (2018)
- Boys – 3.7 percent (36.5/1,000)
- Girls – 0.9 percent 8.8 per 1,000
- Overall – 2.3 % (23.0 per 1,000)
Historical descriptions, diagnostic categories and service records are not directly interchangeable with modern autism surveillance.
Asperger’s disorder was previously a separate diagnosis. Its criteria did not include clinically significant early language delay, but this did not mean an absence of communication difficulties or support needs.
DSM-5 brought autistic disorder, Asperger’s disorder and pervasive developmental disorder not otherwise specified into the autism spectrum disorder category. This classification change alone does not quantify how much prevalence has changed.
Rates of Autism In The US
Let’s have a look at the US autism statistics and autism statistics by state.
CDC’s 2022 ADDM report describes defined surveillance areas, not a ranking of all U.S. states.
Among eight-year-olds in the 2022 ADDM surveillance areas, identified prevalence ranged from 9.7 per 1,000 in Laredo, Texas, to 53.1 per 1,000 at the California site.
Differences in identification practices, available records and access to assessment can affect recorded prevalence. A higher estimate alone does not prove that a state’s services are more effective.
If you reside in a state with a lower prevalence of autism diagnosis, it does not mean that there are fewer autistic persons in your area.
Differences in access to assessment or case finding can contribute to lower recorded prevalence; the estimate alone cannot determine the explanation.
Autism Rates Worldwide: Autism Rates By Country
International estimates differ in age groups, reporting years, diagnostic definitions and case-finding methods.
The evidence presented here does not establish France as the lowest-prevalence country.
Interpreting Higher Recorded Prevalence
A high recorded estimate may reflect more complete identification and should not be interpreted as a reliable worldwide ranking.
Interpreting Lower Recorded Prevalence
WHO reports substantial variation between studies and notes that prevalence remains unknown in many low- and middle-income countries.
Rates Of Autism By Sex
CDC’s 2022 ADDM surveillance found identified autism prevalence among eight-year-olds was 3.4 times as high in boys as girls.
A 2017 systematic review by Loomes and colleagues estimated a male-to-female ratio closer to 3:1 in studies that actively screened populations, suggesting that girls may be underidentified.
That review’s pooled ratio was 4.20 across all included studies and 3.32 in higher-quality studies; estimates varied substantially between studies.
Sex ratios depend on age, population and ascertainment methods; a single ratio should not be treated as universal.
Autism may be overlooked in girls, and an assessment should distinguish autistic characteristics from co-occurring anxiety or depression. Neither condition should automatically be assumed to explain all social or communication difficulties.
Population-screening studies in the 2017 review also identified autism more often in boys, while indicating that diagnosis-only samples can overstate the ratio.
Autism Prevalence By Race And Ethnicity
Race/Ethnicity And The Prevalence Of Autism In Eight-Year-Olds
- CDC 2018 ADDM data, eight-year-olds at 11 U.S. sites: White – 2.12 percent (21.2 per 1,000)
- Black – 2.2 percent (22.3 per 1,000)
- Asian/Pacific Islander – 2.2 percent (22.2 per 1000)
- Hispanic – 2.3 percent (22.5 per 1,000)
- Overall – 2.3 percent (23.0 per 1,000)
According to the CDC, autism is widespread in all races and ethnic groups, but it is occasionally underdiagnosed in particular populations.
Patterns have changed across surveillance years. In 2022 ADDM data, identified prevalence was lower among non-Hispanic White children than among Black, Hispanic, and Asian or Pacific Islander children.
Recorded racial and ethnic differences should not be interpreted as evidence of an inherent biological susceptibility; access to assessment and identification practices affect these comparisons.
More similar overall estimates in some earlier surveillance years did not mean that disparities in assessment or support had disappeared.
There appears to be no evidence that white children are more prone than other children to acquire autism.
CDC reports continuing disparities in timely assessment and identification, emphasizing the need for equitable access to services.
Prevalence Of Autism By Age
A CDC modeling study estimated that 2.21 percent of U.S. adults aged 18–84 had autism in 2017.
The prevalence numbers are often estimated based on children, with the majority of diagnoses given to children.
In the United States, the American Academy of Pediatrics recommends autism-specific screening at both 18 and 24 months. Screening is not a diagnosis; assessment by an experienced professional can often provide a reliable diagnosis by age two.
Discuss developmental concerns with the child’s healthcare professional whenever they arise; do not wait for the next scheduled screening visit.
Autism may be recognized in childhood or much later. Adults may seek assessment because of longstanding difficulties that were previously missed or attributed to another condition.
Early signs can sometimes be detected before age two. Routine screening at 18 and 24 months is distinct from a full diagnostic evaluation, and additional assessment is appropriate whenever concerns arise.
Fluent speech does not rule out autism or eliminate communication difficulties; age at recognition varies.
Autism may be underrecognized in girls, which can delay identification; diagnostic timing varies with the population and individual presentation.
Autism Among Adults
WHO’s global estimate is about 1 in 127 people for 2021. Separately, a CDC study published in 2020 modeled autism prevalence among U.S. adults aged 18–84 at 2.21 percent for 2017.
The modeled 2017 estimate was about 5.4 million U.S. adults aged 18–84, including diagnosed and undiagnosed people.
Employment studies should distinguish having ever held a job, current employment, hours worked and whether work matches the person’s skills and preferences.
Special-education eligibility statistics describe students receiving specified services, not autism prevalence among all students or all disabled children.
Support should reflect the person’s needs and preferences and may include early intervention, educational accommodations, communication support and self-advocacy.
Children who use few or no spoken words need individualized communication assessment and support, which may include augmentative and alternative communication.
Some autistic children are identified early, while others are recognized later. Access to appropriate family, educational and healthcare support is not universal.
Autistic children can have co-occurring ADHD, anxiety or depression. Estimates vary with the population and assessment method, and these conditions should be assessed individually.
Autistic people may have co-occurring anxiety disorders, OCD or bipolar disorder. These are distinct diagnoses; OCD is classified among obsessive-compulsive and related disorders, not anxiety disorders.
Autism can co-occur with genetic conditions such as Fragile X syndrome and Down syndrome. It is a separate diagnosis, not a universal symptom of either condition.
Individuals with autism may also suffer from gastrointestinal difficulties, epilepsy, seizures, and difficulty sleeping.
Motor difficulties can co-occur with autism and may affect activities such as balance or handwriting. Questionnaire screening for motor impairment is not equivalent to a confirmed diagnosis.
Living arrangements and support needs vary among autistic adults; findings from an unspecified sample should not be generalized to all adults.
Adult support planning should respect the person’s own relationship goals, preferences and choices rather than assume a particular marital outcome.
Autism is one of several special-education eligibility categories. An educational classification is not interchangeable with a clinical diagnosis or a population prevalence estimate.
Shattuck and colleagues’ 2012 U.S. study of a special-education cohort aged 19–23 found that 34.7 percent of autistic participants had attended a two- or four-year college during the first six years after high school.
In that study, more than half of those who had left high school within the preceding two years had participated in neither employment nor postsecondary education. These historical findings are not a current estimate for all autistic young adults.
