Developmental intergroup theory posits that when environments make social-group membership salient, children will be particularly likely to apply categorization processes to social groups, thereby increasing stereotypes and prejudices. To test the predicted impact of environmental gender salience, 3- to 5-year-old children (N = 57) completed gender attitude, intergroup bias, and personal preference measures at the beginning and end of a 2-week period during which teachers either did or did not make gender salient. Observations of peer play were also made at both times. After 2 weeks, children in the high- (but not low-) salience condition showed significantly increased gender stereotypes, less positive ratings of other-sex peers, and decreased play with other-sex peers. Children's own activity and occupational preferences, however, remained unaffected.