The aim of this study was to explore the role of ethnic origin in the treatment of acute bipolar mania. Treatment outcomes were studied in a post-hoc analysis of African–American (AA, n=41) and Caucasian (CA, n=190) adults treated with olanzapine in three studies conducted in the United States of America. Baseline demographics were similar except that the AA cohort had fewer women compared with the CA cohort (37 vs. 58%; P=0.01). Daily mean modal olanzapine dose and study discontinuation rate for AA and CA were: 16.2 mg vs. 16.6 mg and 41.5 vs. 25.3% (P=0.03), respectively. There were four (23.5% of discontinuers) and 19 (39.6% of discontinuers, P=0.14) discontinuations because of a poor response in the AA and CA groups, respectively. Drug exposure for the AA cohort was 18.7 days and that of the CA cohort was 19.3 days. Both cohorts showed similar symptom improvements, and safety outcomes were not statistically significantly different except for the following treatment-emergent adverse event frequencies for AA and CA cohorts, respectively: agitation (24.4 vs. 10.5%, P=0.04); dysmenorrhoea (20.0 vs. 3.6%, P=0.04); and dizziness postural (7.3 vs. 1.1%, P=0.04). Although study findings [limited by a smaller (18% of total population) AA cohort] need replication, they suggest that while many outcomes were similar in both cohorts, clinicians could benefit from the awareness of factors in the AA population that possibly influence study discontinuation rates, treatment-emergent adverse event reporting, and participation by sex.