About 1 in 36 children in the United States has autism spectrum disorder, based on the most recent data from the CDC’s surveillance year of 2020. That makes autism one of the most common developmental conditions in the country, affecting roughly 2.8% of 8-year-olds. Among adults, an estimated 1 in 45 has autism, though adult diagnoses are harder to track because systematic screening didn’t exist for earlier generations.
How Prevalence Has Changed Over Time
The CDC’s Autism and Developmental Disabilities Monitoring (ADDM) Network has tracked autism among 8-year-olds every two years since 2000. That first estimate found a prevalence of 1 in 150 children. Over two decades, the number climbed steadily to 1 in 36 by 2020, more than a fourfold increase.
That rise doesn’t necessarily mean autism itself is becoming more common in a biological sense. Much of the increase is explained by broader diagnostic criteria, better awareness among parents and pediatricians, and improved screening in communities that were historically underserved. In the early 2000s, many children with milder traits or those from minority backgrounds were simply never evaluated. As screening expanded, so did the numbers.
That said, researchers haven’t ruled out the possibility that some portion of the increase reflects a genuine rise. Environmental factors, older parental age at conception, and other variables remain active areas of investigation. The honest answer is that both better detection and real increases likely play a role, though the relative contribution of each is still unclear.
Boys Are Diagnosed Far More Often Than Girls
Autism is consistently identified more frequently in boys than in girls. Historically, the ratio has hovered around 4 to 1, meaning for every girl diagnosed, roughly four boys receive the same diagnosis. Recent data suggest that gap may be narrowing slightly as clinicians become better at recognizing how autism presents differently in girls.
Girls with autism are more likely to mask their social difficulties by mimicking peers, which can delay or prevent diagnosis entirely. They also tend to show fewer repetitive behaviors, one of the traits clinicians look for most readily. The result is that many girls aren’t identified until later in childhood, adolescence, or even adulthood, when the demands of social life outpace their ability to compensate.
Racial and Ethnic Gaps Are Closing
For years, white children were diagnosed with autism at significantly higher rates than Black, Hispanic, and other minority children. That gap wasn’t biological. It reflected unequal access to developmental screening, language barriers, and cultural differences in how families and providers interpreted behavior. In many communities, autism traits were attributed to shyness, discipline issues, or language delays rather than a developmental condition warranting evaluation.
More recent CDC data show those demographic patterns in identification have changed substantially. As awareness campaigns reached broader communities and Medicaid coverage expanded for developmental evaluations, diagnosis rates among Black and Hispanic children rose. In some ADDM Network sites, prevalence among Black children now matches or exceeds that of white children, suggesting the earlier disparity was primarily one of access, not occurrence.
When Children Are Typically Identified
The American Academy of Pediatrics recommends autism-specific screening at 18 and 24 months of age during routine well-child visits. This complements the broader developmental screening recommended at 9, 18, and 30 months. The rationale is straightforward: autism can be reliably detected as early as 18 months, and children who receive support during those early years have significantly better outcomes in communication, social skills, and independence.
In practice, many children are diagnosed later than guidelines suggest. The median age of diagnosis in the U.S. has been around 4 to 5 years, depending on the community. Children with more obvious traits, such as limited speech or significant repetitive behaviors, tend to be identified earlier. Children with subtler presentations, average or above-average language skills, or those who are girls or from underserved communities often wait years longer.
The ADDM Network began tracking 4-year-olds in 2010 specifically to measure how well early identification efforts are working. Comparing prevalence at age 4 to prevalence at age 8 in the same community gives a sense of how many children are being caught early versus slipping through and getting diagnosed later in elementary school.
Autism in Adults
An estimated 1 in 45 adults in the U.S. has autism. This figure is less precise than the childhood estimate because no equivalent surveillance system tracks adults. Many adults on the spectrum were never diagnosed as children, either because they grew up before widespread screening existed or because their traits were mild enough to go unrecognized.
A growing number of adults are now seeking diagnosis for the first time, particularly women and people who learned to camouflage their differences. For some, a late diagnosis provides a framework that explains decades of social difficulty, sensory sensitivities, or burnout. It can also open the door to workplace accommodations and therapeutic support that weren’t previously available to them.
The Cost of Late Versus Early Support
Early intervention carries both personal and economic benefits. One widely cited cost-benefit model estimated that children who received intensive early intervention between age 2 and school entry saved between $656,000 and $1,082,000 per person over a lifetime in reduced special education costs, healthcare expenses, and disability services. Intensive behavioral support during preschool years typically runs 20 to 40 hours per week and is expensive upfront, but the long-term savings in reduced need for lifelong support are substantial.
These numbers underscore why screening timelines matter so much. The earlier a child is identified, the sooner they can access the kind of structured support that builds communication and adaptive skills during the period of greatest brain plasticity. Children who start intervention before age 3 consistently show larger gains than those who begin later, even when the intervention itself is identical.
Why the Numbers Keep Rising
If you’re looking at the trajectory from 1 in 150 to 1 in 36 and wondering whether it will keep climbing, the answer is probably yes, at least in the near term. The ADDM Network monitors 16 sites across the country, and prevalence varies meaningfully from one site to another. Communities with robust screening programs and well-funded evaluation services consistently report higher rates, not because autism is more common there, but because they’re better at finding it.
As screening infrastructure improves in more regions, and as diagnostic criteria continue to capture the full range of autistic presentations, the national estimate will likely continue to rise. At some point, the number will plateau once identification catches up with true prevalence. Whether that ceiling is 1 in 30, 1 in 25, or something else, no one can say with certainty yet.