Autism spectrum disorders (ASDs) have long been noted to affect many more males than females, and accordingly, sex-specific factors have been hypothesized to increase males' risk for, or protect females from, ASDs. However, no such factor has been definitively implicated in ASD etiology that can account for its male-biased prevalence, nor is it known whether the major drivers of sex-differential liability act as male-specific risk factors or as female-specific protective factors. Bringing fresh evidence to this question, Robinson et al.(1) report in PNAS the strongest support to date for a female protective effect (FPE) against autistic behavior, as measured quantitatively in general population samples.The premise of the FPE model is simple: if a female-specific factor protects females across the board from reaching the threshold for ASD diagnosis, then those females who are affected are likely carrying a greater etiological load (e.g., genetic variants or environmental influences) than affected males who lack this female-specific protective factor. Because ASDs are highly heritable (2), supporting a role for familial genetic variation in ASD etiology, this model further predicts that relatives of autistic females should be at increased risk for ASD when compared with relatives of autistic males, who carry less genetic liability on average. A complementary model is that of male-specific risk for ASDs, in which a male-specific factor, such as testosterone, increases males' vulnerability to etiological factors relative to females (3, 4). Both models can account for a male predominance among ASD cases, and are not necessarily exclusive. Recent studies of de novo variation in ASD cases appear to support the FPE model, finding a greater proportion of female cases with deleterious, large copy number variants and single nucleotide variants, compared with males (5-11). One of these studies also noted an increased frequency of copy number variants in unaffected female siblings of autistic cases, lending further support to the theory that females are able to withstand greater genetic liability than males without becoming affected (7).In contrast to this evidence from de novo genetic risk, studies addressing heritable risk for ASD in families ascertained for research have been less supportive of the FPE model. For example, a study of over 800 families with an autistic child did not find an increased rate of ASD among first-and second-degree relatives of female ASD probands compared with males (12). Additionally, two prospective surveillance studies of the high-risk infant siblings of earlier-born ASD probands find a statistically significant effect of sibling, but not proband, sex on ASD risk on the recurrence rate for ASDs (13,14). Findings from a recent, large, twin study are similar, with only the sex of the cotwin significantly affecting recurrence rate (2). However, each of these studies evaluated categorical diagnostic status within clinically ascertained populations, and as Robinson et al. (1) aptly emphasize, under...