Despite the monumental follow-up studies of patients with manic-depressive illness by Lund quist (i@@), Rennie (1942), Hastings (1958), and more recently, Shobe (1971), the develop ment of the concept of unipolar and bipolar forms of affective disorders with clinical (Brodie and Leff, 1971), genetic (Dunner et a!., ‘¿ 970; Winokur ci al., 1969), and biologic differences (Buchsbaum et a!., 1971; Cohn et al., i@@'o), has necessitated a revaluation of the question of outcome in this psychiatric illness. The avail ability of lithium carbonate for both acute and prophylactic treatment of mania (Schou, 1968; Coppen ci a!., 1971), and possibly depression (Goodwin et a!., 1972), has also increased the clinical importance of the unipolar-bipolar distinction. The purpose of this study is to provide further information regarding the course of bipolar manic-depressive illness by reporting the level of functioning, recurrence of episodes, and quality of life at follow-up assessed in a group of patients formerly hospitalized for mania at the National Institutes of Health. Previous studies of the course of manic depressive illness which have distinguished uni polar from bipolar patients include those of Lundquist (i@5), Pens (1968), and Winokur (ig6g).* Lundquist (i@@), interviewed 103 manic patients up to 30 years after the onset of illness, and reported 87 (85 per cent) to be ‘¿ socially recovered', although another seven showed ‘¿ mental symptoms of insufficiency'. He concluded that (bipolar) manic-depressive ill ness has a generally good prognosis. Perris reported on 131 bipolar patients for an average of 20 years after the onset of illness, and found * A number of recent studies of the effect of lithium onbipolar patients have been reported; however, these are primarily short-term and concerned mostly with the effect of lithium on preventing recurrent episodes of illness. that bipolar patients had significantly more episodes of illness than 138 unipolar patients followed over the same period. Penis' study does not report levels of functioning between episodes. Winokur (1969) followed 28 bipolar patients for an average of two years following hospitalization for a manic episode and found only four patients to be ‘¿ well in every way'. He concluded that bipolar manic-depressive illness ‘¿ is associated with a high degree of psychiatric and social morbidity'. Several studies of manic-depressive illness which do not differentiate between unipolar and bipolar types nevertheless contain some information on the course of bipolar illness. Renme (1942) studied 208 manic-depressives over 20 years and concluded that ‘¿ cyclothymic [bipolar] cases tend oftener to chronicity [than unipolar cases]'. Hastings (1958) interviewed 42 patients six to 12 years after hospitalization for a manic episode and found only 40 per cent to have ‘¿ a reasonably satisfactory social adjust ment'. Bratfos and Haug (1968) studied 216 patients an average of six years after discharge from hospital and found 19 of 42 manics to have a chronic cour...