Although fluoxetine might be more effective than placebo for treating adolescent depression without major comorbidity, little is known about the response of depressive symptoms to antidepressants in adolescents with comorbid conduct disorder (CD) and substance use disorders (SUD). Male adolescents, who remained or became depressed after > or = 1 month of abstinence from abused substances during residential treatment for SUD, were treated in an open trial for > or = 7 weeks with a fixed dose of 20 mg of fluoxetine. The eight adolescents (ages 14-18 years) with CD, SUD, and major depression were not in drug withdrawal or receiving other pharmacotherapy. A > or = 50% improvement was observed in mean scores on Ten Point Depression Scale rated by clinician (p < 0.01) and patients (p < 0.01), Carroll Self-Ratings for depression (p < 0.02), and Severity of Illness scores on the Clinical Global Impression (p < 0.01). Of the eight adolescents, seven showed marked improvement and wished to continue fluoxetine after the trial. Side effects were mild and transient. No subject required dosage reduction or discontinuation of medication because of side effects. Fluoxetine appeared useful in treating substance-dependent delinquents whose major depressions persisted or emerged after 4 weeks of abstinence. These preliminary findings justify a controlled trial in such youths.