This study examined transactional associations between responses to peer stress and depression in youth. Specifically, it tested the hypotheses that (a) depression would predict fewer effortful responses and more involuntary, dysregulated responses to peer stress over time; and (b) fewer adaptive and more maladaptive responses would predict subsequent depression. Youth (M age=12.41; SD=1.19; 86 girls, 81 boys) and their maternal caregivers completed semi-structured interviews and questionnaires at three annual waves. Multi-group comparison path analyses were conducted to examine sex and stress-level differences in the proposed reciprocal-influence model. In girls and in youth exposed to high levels of peer stress, maladaptive stress responses predicted more depressive symptoms and adaptive stress responses predicted fewer depressive symptoms at each wave. These findings suggest the utility of preventive interventions for depression designed to enhance the quality of girls' stress responses. In boys, depression predicted less adaptive and more maladaptive stress responses, but only at the second wave. These findings suggest that interventions designed to reduce boys' depressive symptoms may help them develop more adaptive stress responses.