There is no single registry tracking every new autism diagnosis in the United States each year, so an exact annual count doesn’t exist. What we do have is prevalence data from the CDC, which estimates that roughly 1 in 31 eight-year-olds (about 32.9 per 1,000) were identified with autism spectrum disorder as of 2022. With approximately 3.6 million children born in the U.S. each year, that translates to an estimated 115,000 or more children per birth cohort who will eventually receive an autism diagnosis. The true annual number of new diagnoses is likely higher, because it also includes teenagers and adults identified later in life.
U.S. Prevalence Among Children
The CDC tracks autism prevalence through its Autism and Developmental Disabilities Monitoring (ADDM) Network, which reports updated estimates every two years based on data from 16 sites across the country. The most recent report, based on 2022 data, found an overall prevalence of roughly 33 per 1,000 among eight-year-olds. That rate varied significantly by race and ethnicity. Asian or Pacific Islander children had the highest recorded prevalence at 38.2 per 1,000, followed by American Indian or Alaska Native children at 37.5, Black children at 36.6, Hispanic children at 33.0, and white children at 27.7.
This is a notable shift from earlier years, when white children were diagnosed at higher rates. The gap has reversed as screening and diagnostic access have expanded in communities that were historically underserved. Black children are now about 1.3 times as likely to be identified as white children, and Asian or Pacific Islander children are about 1.4 times as likely.
How Much Diagnoses Have Increased
Autism diagnoses have increased by roughly 300% over the past 20 years, according to Johns Hopkins Bloomberg School of Public Health. In the early 2000s, the CDC estimated about 1 in 150 children had autism. Today that figure is closer to 1 in 31. Most researchers attribute the bulk of this increase to broader diagnostic criteria, greater awareness among parents and clinicians, and improved screening in historically underdiagnosed populations, rather than a true spike in the underlying condition.
The shift from the older DSM-IV diagnostic manual to the current DSM-5 actually narrowed the criteria slightly. A CDC-funded study found that about 81% of children who qualified under the old definition also met the new one, which would have lowered the estimated prevalence from 11.3 to 10.0 per 1,000 for the 2008 data year. In practice, that predicted dip never materialized because clinicians adapted their documentation and more children were being evaluated in the first place.
Global Estimates
The World Health Organization estimates that about 1 in 127 people worldwide had autism as of 2021. That figure is considerably lower than U.S. estimates, largely because many countries lack the screening infrastructure and diagnostic capacity that drives identification in wealthier nations. Reported prevalence varies enormously from one study to the next depending on the country, the methods used, and the age group examined.
The Gender Gap Is Shrinking
Autism has long been considered far more common in boys, with older estimates citing a 4-to-1 male-to-female ratio. That gap is closing fast. A large population-based study published in The BMJ found that by 2022, the cumulative male-to-female ratio had dropped to about 1.2 to 1 by age 20. For people older than 15 in birth cohorts from 2000 onward, the ratio was at or below 1, meaning women and girls were being diagnosed at the same rate or higher. Projections from the study suggested the ratio could reach full parity by 2024.
At younger ages the gap persists. Among 10-year-olds, boys are still diagnosed about three times as often as girls. This likely reflects the fact that autism presents differently in many girls and women, who are more often identified in adolescence or adulthood once social demands increase and masking strategies become harder to sustain.
Late and Adult Diagnoses
A growing share of autism diagnoses are happening in the teenage years and beyond. Data from a large clinical program at the University of North Carolina found that 15.4% of people in their sample received what they classified as a “late” diagnosis, meaning age 13 or older. That proportion increased with each passing year of the study period, which ran from 2000 to 2021. Women were significantly more likely than men to be diagnosed late, with males having a 38% lower chance of receiving a late diagnosis compared to females.
This trend matters for the annual count question. Many of the new diagnoses being made each year aren’t in young children. They’re in teenagers, college students, and adults who grew up before screening was routine or whose symptoms didn’t match the stereotypical presentation clinicians were trained to spot. No national database tracks these adult diagnoses comprehensively, which means the total number of new diagnoses per year across all ages is almost certainly higher than estimates based on childhood prevalence alone.
Wide Variation by Geography
Autism identification rates differ dramatically depending on where a child lives. Among ADDM Network sites in 2022, the highest recorded prevalence was in California at 53.1 per 1,000 children, while the lowest was in Texas at 14.3 per 1,000. That nearly fourfold difference doesn’t mean autism is more common in California. It reflects differences in how aggressively states screen children, how well-resourced their developmental services are, and how diagnosis data gets collected and reported. The CDC has noted that no research supports the idea that living in a particular community actually increases the chance of having autism.
For families, this means the likelihood of your child being identified depends partly on your zip code. In areas with fewer specialists and longer wait times for evaluations, children who would qualify for a diagnosis in a well-resourced state may go unidentified for years or never receive a formal diagnosis at all.