We present multi-method case studies of two Zimbabwean primary schools - one rural and one small-town. The rural school scored higher than the small-town school on measures of child well-being and school attendance by HIV-affected children. The small-town school had superior facilities, more teachers with higher morale, more specialist HIV/AIDS activities, and an explicit religious ethos. The relatively impoverished rural school was located in a more cohesive community with a more critically conscious, dynamic and networking headmaster. The current emphasis on HIV/AIDS-related teacher training and specialist school-based activities should be supplemented with greater attention to impacts of school leadership and the nature of the school-community interface on the HIV-competence of schools.