The study examined the predictive utility of blame attributions for maltreatment. Integratingtheory and research on blame attribution, it was predicted that self-blame would mediate ormoderate internalizing problems, whereas other-blame would mediate or moderate externalizingproblems. Mediator and moderator models were tested separately. Adolescents (N =160, ages 11–17 years) were randomly selected from the open caseload of a childprotection agency. Participants made global maltreatment severity ratings for each of physicalabuse, psychological abuse, neglect, sexual abuse, and exposure to family violence. Participantsalso completed the Attribution for Maltreatment Interview (AFMI), a structured clinical interviewthat assessed self- and perpetrator blame for each type of maltreatment they experienced. TheAFMI yielded five subscales: self-blaming cognition, self-blaming affect, self-excusing,perpetrator blame, and perpetrator excusing. Caretaker-reported (Child Behavior Checklist) andself-reported (Youth Self Report) internalizing and externalizing were the adjustment criteria.Controlling for maltreatment severity, the AFMI subscales explained significant variance inself-reported adjustment. Self-blaming affect was the most potent attribution, particularly amongfemales. Attributions mediated maltreatment severity for self-reported adjustment but moderatedit for caretaker-reported adjustment. The sophistication and relevance of blame attributions toadjustment are discussed, and implications for research and clinical practice are identified.