The hypothesis is experimentally examined that a positive behavioral repertoire can ameliorate the negative social impact of aggressive behavior. Sixty-six fourth- through sixth-grade girls viewed and rated a series of videotaped interactions between two female confederates. The proportion of positive to aggressive behavior viewed by each subject was manipulated across six experimental conditions by varying the number of positive behaviors exhibited by one of the confederates (target) while the number of aggressive behaviors displayed by her remained constant (i.e., 0%, 20%, 43%, 60%, 80%, and 100% positive). Each subject provided sociometric ratings of the target and selected trait descriptors that best described her from a checklist. The results indicated that positive behavior exhibited in high rates (80% or more, compared to aggressive behavior) offset the negative effects of aggression on both sociometric ratings and positive and negative descriptors.