Gender assignment of children with intersexuality and related conditions has recently become highly controversial. On the basis of extensive animal research and a few human case reports, some authors have proposed the putative masculinization of the brain by prenatal hormones-indicated by the degree of genital masculinization-as the decisive criterion of gender assignment and have derived the recommendation that 46,XX newborns with congenital adrenal hyperplasia (CAH) and full genital masculinization should be assigned to the male gender. The purpose of this study was to test in CAH girls of middle childhood the assumption that prenatal androgens determine the development of gender identity. Fifteen girls with CAH (range of genital Prader stage, 2-4/5), 30 control girls, and 16 control boys (age range, 5-12 years) underwent 2 gender-play observation sessions, and a gender identity interview yielding scales of gender confusion/dysphoria. About half a year earlier, mothers had completed 2 questionnaires concerning their children's gender-related behavior. The results showed that, as expected, CAH girls scored more masculine than control girls on all scales measuring gender-related behavior, with robust effect sizes. By contrast, neither conventionally significant differences nor trends were found on the 3 scales of the gender identity interview. We conclude that prenatal androgenization of 46,XX fetuses leads to marked masculinization of later gender-related behavior, but the absence of any increased gender-identity confusion/dysphoria does not indicate a direct determination of gender identity by prenatal androgens and does not, therefore, support a male gender assignment at birth of the most markedly masculinized girls.