Boys are diagnosed with autism roughly 3.4 times more often than girls, with CDC surveillance data from 2022 showing a prevalence of about 49 per 1,000 among boys compared to 14 per 1,000 among girls. But that ratio has been shrinking, and the reasons behind it turn out to be a mix of genuine biology, genetics, and diagnostic blind spots that have historically missed girls.
The Female Protective Effect
One of the strongest findings in autism genetics is that girls appear to need a larger genetic “push” before they develop the condition. Researchers call this the female protective effect. When scientists compare the DNA of autistic boys and girls, they consistently find that girls carry more high-impact genetic mutations: about 17% of autistic girls carry at least one major risk variant, compared to 11% of autistic boys. In other words, girls seem to tolerate more genetic risk before crossing the threshold into diagnosable autism. Something about female biology raises that threshold, though scientists still don’t know exactly what it is.
This isn’t a small or disputed finding. Multiple large genetic studies, including data from two of the biggest autism research databases (the Simons Simplex Collection and SPARK), consistently support the pattern. It also shows up in family studies: siblings of autistic girls tend to have higher rates of autism than siblings of autistic boys, suggesting the families of diagnosed girls carry a heavier overall genetic load.
X Chromosome Vulnerability
Part of the answer lies in basic chromosome biology. Girls have two X chromosomes; boys have one X and one Y. When a gene on the X chromosome has a harmful mutation, girls often have a backup copy on their second X chromosome that can compensate. Boys don’t have that safety net.
This matters for autism because at least one gene directly tied to autism risk, called NLGN4X, sits on the X chromosome. This gene helps brain cells form and maintain the connections (synapses) they use to communicate. There is a version of the gene on the Y chromosome, but NIH-funded research found the Y version produces a protein that doesn’t reach the cell surface as effectively, making it less capable of building working synapses. So when a boy inherits a mutation in the X-linked version, the Y version can’t pick up the slack. A girl with the same mutation on one X chromosome can often rely on her other copy.
Prenatal Hormones: A Weaker Link Than Expected
A popular theory known as the “extreme male brain” hypothesis proposed that higher prenatal testosterone exposure pushes brain development toward patterns associated with autism, such as strong systemizing tendencies and weaker empathy. Since male fetuses produce more testosterone, this seemed like a tidy explanation for the sex gap.
The evidence has been underwhelming. Studies measuring testosterone in amniotic fluid found some correlations with parent-reported autistic traits in young children, but an independent replication failed to confirm the link. A more recent study tracking children into adolescence found no direct association between prenatal testosterone levels and autistic traits in teenagers of either sex. There were hints of a connection in kids who went through puberty earlier than average, but the finding was statistically fragile. Meanwhile, a brain imaging study that directly tested whether autistic brains look more “male-typical” found that the apparent difference disappeared entirely once researchers accounted for brain size. Prenatal hormones may play some role, but they don’t appear to be a primary driver of the sex gap.
How Autism Looks Different in Girls
Autism was first described in the 1940s based on case studies that were overwhelmingly male: 8 of Leo Kanner’s 11 original cases and all 4 of Hans Asperger’s cases were boys. The diagnostic criteria that evolved from those early descriptions naturally reflect how autism presents in males. This matters because autistic girls and women often look different in clinical settings.
Girls with autism tend to show fewer of the repetitive, restricted behaviors that are a hallmark of diagnosis, and the interests they do develop intensely may appear more socially typical (animals, celebrities, fiction) rather than the stereotypically “autistic” interests like train schedules or electronics. Their social difficulties are also more likely to manifest as internalizing problems like anxiety and withdrawal rather than the externalizing, disruptive behaviors that prompt parents and teachers to seek evaluation for boys.
Perhaps most importantly, many autistic girls develop sophisticated strategies for masking their difficulties. Researchers call this camouflaging, and it involves consciously copying the social behaviors of peers, rehearsing conversations, and forcing eye contact or facial expressions that don’t come naturally. Studies using standardized observation tools have found that autistic women display more vocal expressiveness than autistic men and show stronger surface-level social engagement, even when their underlying social cognition is equally impaired. The result is that their autism is harder to detect in a clinical assessment designed around male presentation patterns.
The Gender Gap Is Narrowing
The male-to-female ratio has been dropping steadily. CDC data shows it went from 4.2 in 2018 to 3.8 in 2020 to 3.4 in 2022 among 8-year-olds. But the trend is even more dramatic in older age groups. A large population-based study published in The BMJ in 2025 tracked autism diagnoses across birth cohorts and found that among people diagnosed after age 15, the ratio was no longer above 1 in the most recent data period (2020 to 2022). By age 20, the cumulative ratio had fallen to about 1.2 in 2022, and projections suggested it could reach parity by 2024.
This pattern, where the ratio stays around 2 to 4 for young children but narrows dramatically in teenagers and adults, strongly suggests that many girls and women are being identified later rather than not being autistic at all. In one study from Uganda, girls were diagnosed at an average age of 7.8 years compared to 6.6 years for boys, and about 25% of girls weren’t diagnosed until after age 10, compared to 17% of boys. That delay adds up across populations.
The CDC has also noted that interpreting the narrowing ratio requires caution. While the ratio itself is shrinking, the absolute gap in prevalence between boys and girls has actually widened, from 27.7 per 1,000 in 2018 to 34.9 per 1,000 in 2022, because overall autism identification is rising faster in boys in raw numbers. The picture is complex: more girls are being found, but the identification of boys is accelerating too.
Biology and Bias Both Matter
The honest answer to why autism is more common in boys is that it’s partially both a real biological difference and a measurement problem. The female protective effect is well supported. Girls genuinely do seem to need more genetic disruption before developing autism, and X chromosome biology gives them a buffer that boys lack. At the same time, decades of male-centered diagnostic criteria, camouflaging behaviors in girls, and a clinical culture more attuned to recognizing autism in boys have created a systematic undercount of autistic females. The rapidly narrowing ratio in older age groups suggests that as clinicians get better at recognizing autism in women and girls, the gap will continue to close, though it likely won’t disappear entirely.