Sex differences in play are apparent in a number of mammalian species, including humans. Prenatal testosterone may contribute to these differences. The authors report the first attempt to correlate gender-typed play in a normative sample of humans with measurements of amniotic testosterone (aT). Testosterone was measured in the amniotic fluid of 53 children (31 boys and 22 girls). A strong sex difference was observed in aT and, at ages 4.75 to 5.8 years, on a modified version of the Child Game Participation Questionnaire. Hierarchical regression analyses on the entire group and within-sex correlations suggested that variations in aT did not contribute to individual differences in game participation as reported by the mother. A critique of explanations for this finding is presented.