The complete population of adolescents in a residential and day-treatment program over a 4-year period, 532 youths, served in two studies. Along with residential and day-treatment settings, predictive variables of interest were the number of hours spent in group, individual, and family therapy. A total of 227 adolescents qualified for Study 1 which found a reduction of rates of criminal charges from pre- to posttreatment. Study 1 also found that hours in group therapy explained the most variance in the reduction in rates of criminal charges, followed closely by hours in individual therapy. Hours in family therapy was not a significant predictor. A total of 430 adolescents qualified for Study 2, which found that residential treatment was associated with greater reductions in adult correctional commitments than day treatment. Implications stress the need for further research examining the relationships between therapeutic components of residential treatment and behavioral outcomes.