This study examined the efficacy of the Juvenile Justice Anger Management (JJAM) Treatment for Girls, an anger management and aggression reduction treatment designed to meet the unique needs of adolescent girls in residential juvenile justice facilities. This randomized controlled trial of JJAM compared changes in levels of anger and aggression among girls who participated in the JJAM treatment with those of girls who participated in treatment as usual (TAU) at the facilities. This study also investigated the theoretical model underlying the JJAM treatment, which proposed that reductions in hostile attribution biases, development of emotion regulation skills, and improvement in social problem solving would serve as mechanisms of action in JJAM. Participants were 70 female youth who ranged in age from 14 to 20 years (M = 17.45, SD = 1.24) and were placed at 1 of 3 participating juvenile justice facilities; 57 youth completed the study and were included in analyses. Results revealed greater reductions in anger, reactive physical aggression, and reactive relational aggression among girls in the JJAM treatment condition when compared to girls in the TAU control condition. The proposed theoretical model was partially supported via significant mediation findings; changes in hostile attribution bias were identified as a significant mechanism of action in the JJAM treatment. Results suggest that JJAM is a promising treatment to effectively reduce anger and reactive aggression among adolescent girls in juvenile justice placements.