In a more sophisticated replication of an earlier study (Abrams and Taylor 1974), we examined 77 manic patients, of whom 29 had never suffered a depressive illness, and had two or more manic attacks. These unipolar manics were similar to the 48 bipolar manics for a wide variety of clinical, phenomenological, historical, laboratory and demographic variables, generally supporting our earlier findings. However, the present sample showed a striking excess of males among the unipolar manics, as well as an increased morbid risk for unipolar depression in first-degree relatives. Although not readily explainable, these differences suggest that it is premature to equate unipolar mania with classical bipolar illness. Further studies of unipolar mania are in progress.