One in 44. One in 36. Now one in 31. That is the run of estimates the CDC has published for autism in eight-year-olds, for 2018, 2020 and 2022 respectively, and the last of them came out on a Tuesday in April 2025.
The figure does not describe the whole country, and the CDC says so itself. It comes from the agency’s Autism and Developmental Disabilities Monitoring Network, which in this round pulled health and school records from 16 sites. Within that spread the range is wide: autism prevalence was under 1% in Laredo, Texas, and 5.3% in California. Averaging those into a national headline is, as ever, a convenience rather than a measurement.
What actually rose
The earlier estimates were 1 in 150 in the 2000s, 1 in 44 for 2018 and 1 in 36 for 2020, according to the AP. As late as the early 1990s the diagnosis went to roughly 1 in 10,000 children, back when it described only severe cases. Then the term widened to cover a group of milder, related conditions, and the counts ballooned.
So the detail that matters is what is being counted. The network does not run a clinical screening of every child. It reads records, and it looks at eight-year-olds because most cases have been diagnosed by that age. Experts told the AP the CDC’s estimate is the most rigorous available and the gold standard.
Health officials largely put the climb down to better recognition through wider screening and better diagnosis. There is no blood test or other biological marker for autism, the AP noted; clinicians judge behaviour. Alongside that, the AP pointed to a surge in therapies and services aimed at children.
The gap between boys and girls
Boys are still identified far more often than girls, and the CDC does not claim otherwise. Across the network the ratio among eight-year-olds was 3.4 boys to each girl. Among four-year-olds it fell to 2.8. The younger the child, the narrower the gap.
The figures from Maryland, where Johns Hopkins researchers run one of the sites, show the same drift. There the number of eight-year-old girls identified with autism passed 1%, or 1 in 93, for the first time. The 2020 report had already put girls above 1% across the network.
“Years ago, we only knew that autism was impacting white males,” said Elise Pas, a research professor at Johns Hopkins. The newest report, she said, shows that it affects everyone, females and children of all racial and ethnic backgrounds included. The racial pattern has indeed turned: the highest rates in this round were among Asian and Pacific Islander, American Indian and Alaska Native, and Black children.
Earlier diagnosis, and the politics
Four-year-olds in the network were nearly twice as likely to have been diagnosed by that age as eight-year-olds had been when they were four. Babies born in 2018 are being identified sooner than those born in 2014.
Health Secretary Robert F. Kennedy Jr. seized on the numbers the next day. He called the rise an “epidemic” and promised to find the “environmental toxin” he said was to blame, with answers expected by September. Researchers pushed back. Catherine Lord, a professor at the David Geffen School of Medicine at UCLA, wrote that it is very unlikely autism is caused by one thing, and many autism specialists say the rise tracks better screening. Studies by the CDC and others ruled out childhood vaccines as a cause roughly two decades ago.
Sources
- Autism rates in US rise again to 1 in 31 kids, CDC says (AP via Scripps News)
- Autism Affects 1 in 31 Children Aged 8 (Neuroscience News / Johns Hopkins)
- RFK Jr. calls rising autism rates an epidemic (NPR)
