This study evaluates implementation levels (dosage) and impact on recidivism of Dialectical Behavior Therapy, as adapted for juveniles in secure residential facilities in Washington State. Actual dosage was compared with agency standards for three DBT modes, coaching on the floor (environmental adherence), individual counseling sessions, and skills group sessions. Recidivism was measured as a conviction for any offense, a felony offense, or a misdemeanor offense committed within 18 months of release from a residential facility. The results of logistic regression analysis, from a sample of 1,031 youth, indicate that increases in individual counseling and skills group sessions did not have a significant effect on recidivism. Increases in environmental adherence, however, were significantly related to reductions in recidivism, the impact being greatest for younger youth and those with mental health risk. Overall, the more that living unit environments were structured for skill generalization, behavioral change, and goal achievement, the better the anticipated outcomes. The challenges related to quality assurance and implementation in secure facilities are discussed.