To assess whether temperament moderates response to treatment for substance use disorders (SUD), n = 88 consecutively referred adolescents with SUD were randomized to cognitive-behavioral (CBT) or psychoeducational (PET) therapies. Principal components analyses reduced the 10-attribute Dimensions of Temperament Revised (DOTS-R) to three factors (Rhythmicity, Activity/Flexibility, and Mood/Sociability). Three response domains (Alcohol Use, Other Substance Use, and Psychological Function) from the Teen Addiction Severity Index (T-ASI) were assessed at baseline and at 3-month post-treatment completion. There was a significant moderator effect on Treatment only for Rhythmicity on T-ASI Alcohol and Psychological domains. For Alcohol use, for youth with poor levels of Rhythmicity, only PET effected significant improvement, whereas for youth with good levels of Rhythmicity, no treatment differences emerged. For Psychological function, for youth with good levels of Rhythmicity, only CBT effected significant improvement, whereas for youth with poor levels of Rhythmicity, no treatment differences emerged. Implications of these findings are discussed. (Contains 2 tables.)