Eligibility for the Nokomis program is limited to young people 14 years of age and older who have not been convicted of a Class I offense or arson, who do not have suicidal tendencies or developmental-emotional disabilities, and who do not have a medical condition that would limit their participation in outdoor challenge activities. The program was evaluated in terms of its effectiveness in meeting three goals: (1) increase youth participation in school and legitimate employment and improve family functioning; (2) suppress and detect criminal activity and drug use while youth are in the community under intensive supervision and comparison youth are still in residential placement; and (3) reduce criminal activity and drug use after youth have completed the program. The evaluation began in 1990 and included 102 young people who were assigned to the program by Michigan juvenile courts between January and November 1991 and a comparison sample of 97 similar young people who were not placed in the program. The evaluation was based on outcome measures of social adjustment--coping skills, family functioning, and self- esteem. Differences between groups in recidivism rates were also assessed. Evaluation findings showed the Nokomis program was significantly more cost-effective than residential placement but produced few outcome differences. Young people in both program and comparison groups made positive gains in social adjustment, but these gains disappeared by the end of the follow-up period. Despite community aftercare, program youth failed more quickly after release to the community. Results suggest the intensive aftercare component of the Nokomis program must be strengthened to derive benefits from short-term alternative placement and to keep youth from relapsing. Policy implications of the evaluation findings are examined. 44 references, 23 notes, 11 tables, and 1 figure