About 1 in 127 people worldwide have autism, which works out to roughly 0.8% of the global population. That figure, based on 2021 data from the Global Burden of Disease Study, translates to an estimated 61.8 million people. In the United States, the numbers run considerably higher, reflecting both broader diagnostic practices and greater access to evaluation services.
Prevalence in US Children
The CDC’s most recent surveillance data, collected in 2022, found that 1 in 31 eight-year-olds had been identified with autism. That’s about 3.2% of children at that age. The number has climbed steadily over the past two decades: in 2000, the estimate was 1 in 150. By 2020, it had reached 1 in 36. Between 2020 and 2022 alone, prevalence jumped by about 22% across the sites tracking it.
These numbers come from the CDC’s Autism and Developmental Disabilities Monitoring Network, which reviews health and education records for children in 16 communities across the country. The figures vary significantly from site to site, largely because communities differ in how many diagnostic services they offer and how aggressively they screen young children.
Prevalence in US Adults
A CDC study using 2017 data estimated that 2.21% of American adults are on the autism spectrum. By population, the states with the largest number of autistic adults were California (about 702,000), Texas (about 450,000), New York (about 342,000), and Florida (about 329,000). Adult prevalence numbers are harder to pin down because many adults, particularly those with fewer support needs, were never evaluated as children and may remain undiagnosed.
Why the Numbers Keep Rising
The sharp increase over the past two decades is driven primarily by changes in who gets identified, not by a sudden biological shift. Several factors overlap.
Broader awareness among parents, teachers, and pediatricians means more children are referred for evaluation at younger ages. Communities that have invested in early screening consistently report higher prevalence, while areas with fewer services report lower numbers. The gap between well-resourced and underserved communities accounts for much of the variation across states and regions.
Historically, autism was diagnosed far more often in white children and in higher-income families. That pattern has reversed in recent years. Black, Hispanic, and lower-income children are now being identified at rates closer to (and in some cases higher than) their white peers. This shift reflects expanded access to diagnostic services rather than a true change in how common autism is among different groups.
Changes to the diagnostic criteria have also played a role, though in a more complicated way. When the DSM-5 replaced the older diagnostic manual in 2013, it merged several previously separate diagnoses (including Asperger’s syndrome) into a single autism spectrum. Some researchers initially worried the new criteria were more restrictive and would reduce diagnoses, and early studies confirmed a modest decrease. Over time, though, clinicians have adapted, and the gap between old and new criteria has narrowed.
The Gender Gap Is Shrinking
Autism has long been described as roughly three to four times more common in males than females. That ratio is changing quickly. A large birth cohort study published in The BMJ found that the male-to-female ratio holds at about 3 to 1 for children diagnosed before age 10, but drops significantly for those diagnosed later. Among people diagnosed between 2020 and 2022, the ratio was no longer above 1 for individuals older than 15. By age 20, the cumulative ratio had fallen to 1.2 males for every female, and projections suggested it would reach parity by 2024.
Several factors explain the historical undercount of autistic women and girls. Girls tend to have stronger social and communication skills on the surface, which can mask the traits clinicians look for. They are also more likely to camouflage, picking up behavioral cues from peers and mimicking speech patterns or facial expressions. Women are more likely than men to receive a psychiatric diagnosis like anxiety or depression before anyone considers autism, a pattern known as diagnostic overshadowing. There has also been concern that the diagnostic tools themselves were developed and validated primarily on male populations, creating a built-in bias.
Global Variation
The worldwide average of 1 in 127 masks enormous differences between countries. The global age-standardized prevalence is about 788 per 100,000 people, but this breaks down unevenly: roughly 1,065 per 100,000 males and 508 per 100,000 females. Countries with well-funded healthcare systems and established screening programs consistently report higher rates, not because autism is more common there, but because they have the infrastructure to detect it.
In lower-income countries, many autistic people are never formally diagnosed. Limited access to trained clinicians, cultural differences in how developmental differences are understood, and competing public health priorities all contribute to underreporting. The WHO notes that reported prevalence varies substantially across studies, and the true global figure is almost certainly higher than current estimates suggest.
What These Numbers Actually Mean
Prevalence figures measure how many people have been identified with autism, not how many people “have” it in some absolute biological sense. The steady climb in numbers over the past 20 years largely reflects a healthcare system getting better at finding people who were always there. Broader criteria, better-trained clinicians, expanded screening in underserved communities, and growing recognition that autism presents differently in girls and women have all contributed to a more complete picture. The 1-in-31 figure for US children and the 1-in-127 global estimate represent the current best count, with the understanding that both are likely underestimates.