Social resources are considered important protectors in traumatic conditions, but few studies have analyzed their role in psychosocial interventions among war-affected children. We examined (1) whether a psychosocial intervention (teaching recovery techniques, TRT) is effective in improving peer and sibling relations, and (2) whether these potentially improved relations mediate the intervention's impacts on children's mental health. Participants were 428 Palestinian children [10–13 (mean = 11.29, standard deviation SD = .68)-year-old girls (49.4 percent) and boys (50.6 percent)], who were cluster-randomized into the TRT and wait-list control groups. They reported the quality of peer (friendship and loneliness) and sibling (intimacy, warmth, conflict, and rivalry) relations, and posttraumatic stress, depressive and psychological distress symptoms, as well as psychosocial well-being at baseline (T1), postintervention (T2), and six month follow-up (T3). Results showed gender-specific TRT intervention effects: Loneliness in peer relations reduced among boys and sibling rivalry reduced among girls. The TRT prevented the increase in sibling conflict that happened in the control group. The mediating hypothesis was partially substantiated for improved peer relations, and beneficial changes in sibling relations were generally associated with improved mental health.