Data from a published double-blind randomized trial comparing olanzapine versus haloperidol in acute mania were used to address the response and tolerability of Latin American patients. Primary efficacy end point was the remission rate (Young Mania Rating Scale score <or=12 and Hamilton Depression Rating Scale score of <or=8). Patients were analyzed on an intent-to-treat basis. The mean modal doses (milligrams per day) were similar in Latin American (OL) (14.2; n = 51) and white (OC) (15.1; n = 120) patients treated with olanzapine, and in Latin American (HL) (7.1; n = 48) and white (HC) (8.5; n = 113) patients treated with haloperidol. At week 6, remission rates were similar among the OL and HL patients (64.7% vs. 68.8%) but were higher in the OC than in HC (49.2% vs. 32.7%; P = 0.012). Significantly more HL than OL patients experienced extrapyramidal symptoms such as akathisia and tremor. Tremor was significantly higher in HL than in HC patients, whereas a significant increase in the Barnes Akathisia Scale and Abnormal Involuntary Movement Scale scores was observed in HC versus HL. Somnolence and weight gain were significantly higher in OL than in OC patients, and more OL and OC patients experienced weight gain in comparison with the HL and HC groups, respectively. The incidence of nonfasting glucose levels above normal levels did not statistically differ between groups. In conclusion, in contrast to our findings among white patients, the Latin American patients who have acute mania did not differ in overall response to olanzapine or haloperidol. The pattern of adverse events differed between treatment groups. Prospective clinical trials in Latin American bipolar populations are justified.