The goal of this 3-year follow-up evaluation was to determine whether the decreases in drinking and in depressive symptoms that were noted during our acute phase study with fluoxetine in comorbid adolescents persisted at a 3-year follow-up evaluation. At the 3-year follow-up evaluation, the group continued to demonstrate significantly fewer DSM criteria for an AUD and fewer BDI depressive symptoms and also consumed fewer standard drinks than they had demonstrated at the baseline of the acute phase study. However, 7 of the 10 participants demonstrated MDD at the 3-year follow-up assessment, and 4 demonstrated an AUD. The presence of a MDD was significantly correlated with the presence of an AUD at both the 1-year and the 3-year follow-up assessments. Four of the participants restarted SSRI medications during the follow-up period. Half of the subjects graduated from college during the 3-year assessment period, despite their residual depressive symptoms and drinking. We conclude that the long-term therapeutic effects of an acute phase trial of fluoxetine plus psychotherapy slowly decrease but did not disappear when fluoxetine is discontinued shortly after the acute phase trial. The high rate of MDD at follow-up suggests that longer term antidepressant medication treatment may be needed for at least some comorbid adolescents.