We examined the clinical and research records of 29 acutely ill hospitalized patients with affective disorder who received only lithium carbonate during their first week of treatment. Nineteen patients (Group I) could be continued on lithium ion alone, while 10 patients (Group 2) needed additional somatic treatment. Compared with Group 1, Group 2 patients were significantly younger at illness onset, more severely ill on admission, clinically more "colorful" in dress and behavior, stayed more than twice as long in the hospital, and (although not statistically significant) had more than twice the morbidity risk for affective disorder in first-degree relatives. At discharge, both groups were equally improved, and 70% of Group 2 patients were receiving lithium alone. We did not confirm previous reports that nonresponders to lithium alone (Group 2) were more overactive or paranoid--destructive or less euphoric--grandiose than responders to lithium alone (Group 1). Our Group 2 patients had a more severe or penetrant form of illness than our Group 1 patients, requiring neuroleptic drugs or ECT in addition to lithium therapy. Eventually, however, they had a satisfactory outcome, suggesting therapeutic optimism and tenacity even in those patients who initially fail lithium alone and require polytreatment.