OBJECTIVE: Stress resilience was assessed in 81 school-age children, from within 32 families acknowledging severe psychosocial stress. Resilient and nonresilient children, identified via competence and behavior disorder measures from school and home, were compared. METHOD: Parents and teachers completed questionnaires and rating scales, and children were observed and assessed at home on attributes of temperament, self-esteem, ability, gender, and mother-child warmth. RESULTS: Individual differences in child and family attributes that were predictive of competent child functioning varied according to the outcome measure used. Teachers' ratings of positive temperament (low emotional reactivity, high social engagement) best discriminated children showing resilience on all indicators, i.e., behavioral and social competence both at home and at school, with maternal warmth and the number of adverse life events the child had recently experienced also contributing. Level of maternal stress and individual differences in child intelligence were related to academic adjustment, but the child's age, sex, ability, and self-concept were not significant discriminators of behavioral adjustment. CONCLUSIONS: The findings emphasize the salience of a positive temperament as a resilience factor as well as the need to consider different estimates and contexts in assessing resilience for children growing up in stressful situations.