CREST treatment methods, oriented toward attitudinal and thought processes rather than social and institutional norms, used reality therapy as well as gestalt techniques and occasional positive reinforcement therapy. Graduate student counselors, assisted by paraprofessional undergraduate students, provided individual and sometimes group and family counseling for each of the 20 or more clients on the team's caseload. Despite differences in counselors, subjects, and mode of measurement, the overwhelming result of four outcome studies of the CREST project has been that CREST-treated groups committed at least 50 percent fewer criminal acts than did clients in other Youth Services Programs. Long-term followups show that this effect is stable over time and that a large proportion of CREST clients maintain a clear record for the duration of the observation period following termination of treatment. In the three areas of academic adjustment studied (absenteeism, grades, and suspensions), the CREST groups exhibited consistent improvement. Also, measures of social adjustment and self-esteem showed that in the first year CREST subjects improved in their ability to handle anger, relate to others, and see themselves as less deviant. Of the CREST treatment subjects 20 percent were found to have clear records during the 2-year followup period, compared to 48 percent of the controls. It is concluded that the dual treatment approach of the CREST model is a successful departure from traditional juvenile offenders rehabilitation models. Two graphs, five data tables and seven references are provided.