Positive friendships have been related to decreasing levels of children's physical aggression over time. While this evidence calls for interventions aimed at helping children build good-quality friendships, tests of causality through experimental manipulations are still needed. The goal of this study was to examine whether an intervention aimed to increase dyadic friendship quality could decrease children's physical aggression at the peer-group level over a school year. A total of 34 aggressive children and each child's best friend were randomly assigned to two groups: an experimental condition where the members in each dyad participated together in 12 weekly intervention sessions and a no-intervention control condition. Multiple sources were used to evaluate changes in friendship quality and children's physical aggression. Results showed an indirect effect of the intervention on decreasing levels of physical aggression through the improvement of one specific feature of friendship quality: conflict resolution. These results point to the usefulness of including this relational aspect in intervention programs targeting aggressive children.