In this 2-year longitudinal study, we hypothesized that sex of the human child ( Homo sapiens ), differences in physical activity, and time of the year would interact to influence preschool children's sex segregation. We also hypothesized that activity would differentially relate to peer rejection for boys and girls. Consistent with the first hypothesis, high-activity girls started off as the most integrated group but became more segregated with time, whereas high-activity boys remained the most segregated group across the duration of the study. The second hypothesis was also supported: For girls only, activity was significantly related to peer rejection during Year 1 only, the time when high-activity girls also interacted frequently with boys. Results are discussed in terms of sexual selection theory and gender boundary violations.