We first proposed the prenatal sex steroid theory of autism 25 years ago to account for a number of then-unexplained observations around autism, including (1) the more frequent diagnosis of autism in male than in female individuals and (2) apparent ‘male-type’ shifts in cognitive traits associated with autism, such as empathizing and systemizing. Here we review 25 years of research testing this theory. Early studies found that higher prenatal testosterone levels were associated with slower social, language and empathy development, greater attention to detail, stronger systemizing and more autistic traits. Subsequent studies suggested that both prenatal androgens and oestrogens are associated with autism. New methods in genetics and using stem-cell-derived neural organoids have further indicated the importance of sex steroid hormones for neurodevelopment, as well as atypical patterns in autism. These new findings support and open new lines of research into the prenatal sex steroid theory of autism. Twenty-five years after the prenatal sex steroid theory of autism was first proposed, the authors of the original theory discuss the current state of evidence and new lines of research.
Baron-Cohen et al. (2026) studied this question.