This report details the long-term course for systematically rediagnosed (and largely chronically ill) patients with schizophrenia (n = 163) and with bipolar (n = 19) and unipolar (n = 44) affective disorders from the Chestnut Lodge, Rockville, Md, follow-up study. Their conditions were assessed and they are described rigorously from multiple outcome perspectives. Except in the realm of symptomatic diathesis, striking differences emerged between these major axis I disorders consonant with Kraepelin's original observations. Roughly two thirds of the schizophrenic patients were functioning marginally or worse at follow-up, compared with one third of the unipolar cohort. The reverse held for better outcomes. Outcome varied little as a function of follow-up interval (time) across all diagnostic categories. Representative case examples serve to place the ratings in meaningful clinical contexts.