This study evaluated the effectiveness of a military-style residential treatment program for adolescents with academic and conduct problems. Two hundred twelve referred adolescents were separated into 3 groups for analyses: (a) adolescents who completed the 22-week program, (b) adolescents who prematurely withdrew, and (c) wait-list controls. Adolescents' socioemotional and behavioral functioning were measured at baseline and 6 months after treatment. Results showed statistically and clinically significant reductions in externalizing symptoms and increases in adaptive behavior associated with treatment. Treatment was also associated with increased likelihood of high school completion or employment and decreased likelihood of alcohol or drug problems and arrest. The relation between treatment participation and outcomes was moderated by adolescents' living environments after treatment, but it was not moderated by age of symptom onset. The benefits of treatment may be partially attributable to the voluntary nature of the intervention.