Examined efficacy of an empirically based intervention using 70 divorced mothers who participated in a 12-session program or a wait-list condition. The program targeted five putative mediators: quality of the mother-child relationship, discipline, negative divorce events, contact with fathers, and support from nonparental adults. Posttest comparisons showed higher quality mother-child relationships and discipline, fewer negative divorce events, and better mental health outcomes for program participants than controls. More positive program effects occurred for mothers' than children's reports of variables and for families with poorest initial levels of functioning. Analyses indicated that improvement in the mother-child relationship partially mediated the effects of the program on mental health.