The past 20 years have seen a shift in the way psychiatrists conceptualize depressive disorders. This evolution in our thinking has been stimulated by the results of epidemiological studies and of long-term clinical follow-up studies, investigations of spectrum disorders, advances in biological psychiatry, and results of large clinically focused national trials. In the past, major depressive disorder was viewed as episodic and self-limited. However, longitudinal observations of patients treated for major depressive disorder show that after recovery from an index episode of major depression, patients remain symptomatic about 60% of the remainder of their lives with another episode of major depression, minor depressive disorder, or subsyndromal symptoms of depression (Judd et al. 1998a). Data from the National Institute of Mental Health (NIMH) Collaborative Depression Study demonstrated that the recurrence rates of major depressive disorder are extraordinarily high over time: 60% at 5 years of follow-up, 75% at 10 years, and 87% at 15 years (Keller et al. 1982, 1984, 1992; Kupfer 1991).