Public healthCDC ADDM Network2018–2022

The ratio is shrinking. The gap is growing. Both are true.

Every CDC autism report delivers the same reassuring headline: the sex ratio in diagnosis keeps narrowing. Boys outnumbered girls 4.2 to 1 in 2018. By 2022 it was down to 3.4 to 1. But count actual children instead of ratios, and the story reverses — the gap between how many boys and girls get diagnosed has never been bigger.

By Kalaivani Chandramohan · August 14, 2026

Autism diagnosis rate per 1,000 8-year-olds, by sex. 2020 figures (43.0 / 11.4) are exactly as CDC reported them; 2018 and 2022 figures are calculated from CDC's published ratio and gap statistics for those years, since the exact split wasn't published in the same report.

The paradox

A ratio and a gap answer different questions

A ratio asks: for every girl diagnosed, how many boys are? That number is falling, because girls' diagnosis rate is climbing faster than boys' — in percentage terms.

A gap asks: how many more boys are diagnosed than girls, in absolute terms? That number keeps growing, because boys started from a much bigger base. A 36% increase on 36 children per 1,000 is a lot more actual children than a 67% increase on 9.

+67%Girls' rate, 2018 → 2022
+36%Boys' rate, 2018 → 2022
+7.2Gap growth, per 1,000

Mechanism

Why boys get found first

This isn't new, and it isn't only an American pattern.

A 2017 meta-analysis pooling 54 studies and nearly 14 million children found that when researchers looked only at kids who already carried a clinical autism diagnosis, the male-to-female ratio was 4.6 to 1. But when they screened whole populations for autistic traits — regardless of who'd already been diagnosed — the ratio dropped to 3.25 to 1.

Male-to-female ratio in autism, by study design. Loomes, Hull & Mandy, Journal of the American Academy of Child & Adolescent Psychiatry (2017).

The authors' conclusion: the widely quoted 4-to-1 ratio isn't really a fact about autism. It's a fact about who gets referred, evaluated, and believed. Girls who meet the same clinical criteria as boys are measurably less likely to walk away with a diagnosis — partly because autism's diagnostic tools were built and validated mostly on male-typical presentations, and partly because girls are more likely to "camouflage" traits socially in ways that read, to a teacher or a pediatrician, as just being quiet.

Look again, later

Wait until they're older, and the picture changes again

The younger you look, the more lopsided autism appears.

CDC's 2022 report found a 2.8-to-1 ratio among 4-year-olds, widening to 3.4-to-1 by age 8. A Swedish national registry study that followed 2.7 million children all the way into adulthood found the ratio kept dropping with age — down to roughly 1.2 to 1 by age 20, with researchers projecting it would reach near-parity around 2024.

Male-to-female diagnosis ratio by age. Ages 4 and 8: CDC ADDM Network, 2022 survey year (United States). Age 20: national birth-cohort registry, Sweden, cumulative incidence through 2022 — a different country and dataset, shown for direction, not direct comparison.

That's a different country and a different dataset — but the direction matches. Many of the "missing" autistic girls of childhood aren't missing. They're just found later, often well past the early-intervention years that make the earliest difference.

Who gets found first, within girls

The girls who do get diagnosed early look the most like the stereotype

Among children who are diagnosed with autism by age 8, girls are still more likely than boys to also carry a co-occurring intellectual disability.

40%Girls with co-occurring ID
32%Boys with co-occurring ID
Share of children with ASD who also have an intellectual disability (IQ ≤ 70), by sex. CDC ADDM Network, 2016 survey year.

It's the same shape as the racial gap in autism diagnosis: the system is quickest to catch the cases that are hardest to miss. A boy with mild, sub-clinical-seeming traits gets referred. A girl with the identical profile — quieter, more able to mimic her peers, less disruptive in a classroom — often doesn't, until something else forces the question.

The narrowing ratio gets reported as progress, and in one sense it is — more girls are being found than ever. But "the ratio is narrowing" and "the number of undiagnosed autistic girls is shrinking" are not the same sentence. Until the gap closes too, the more honest way to read these numbers isn't that girls are catching up. It's that they're still waiting.