OBJECTIVE: This analysis assessed the incidence, severity, onset, and duration of nausea among patients with major depressive disorder (MDD) treated with the new antidepressant duloxetine. METHODS: Data were pooled from 8 double-blind, randomized, placebo- and active comparator-controlled trials employing patients with MDD that were submitted to the US Food and Drug Administration to support duloxetine's new drug application for treatment of MDD. RESULTS: The numbers of patients receiving each regimen were as follows: placebo, n = 777; duloxetine 40 mg/d, n = 177; duloxetine 60 mg/d, n = 251; duloxetine 80 mg/d, n = 363; duloxetine 120 mg/d, n = 348; paroxetine 20 mg/d, n = 359; and fluoxetine 20 mg/d, n = 70. In acute placebo-controlled trials of duloxetine 40 to 120 mg/d, treatment-emergent nausea was reported by more duloxetine-treated patients than those receiving placebo (19.9% [227/1139] vs 6.9% [154/777], respectively; P <0.001). Among duloxetine-treated patients, the median time to onset of nausea was 1 day, and the median duration of nausea was 7 days. The incidence of nausea was similar to placebo rates after 1 week. In paroxetine-controlled studies, the incidence of treatment-emergent nausea in patients receiving duloxetine did not differ significantly from paroxetine (14.4% vs 12.0%, respectively). In head-to-head studies, the incidence of treatment-emergent nausea with duloxetine did not differ significantly from that with fluoxetine (17.1% vs 15.7%, respectively). Most duloxetine-treated patients reported nausea to be mild (52.9%) or moderate (41.4%). Treatment discontinuation secondary to nausea occurred in more duloxetine-treated patients than those receiving placebo (1.4% [16/1139] vs 0.1% [1/777], respectively; P = 0.002). Following abrupt discontinuation after acute treatment, 5.9% of duloxetine-treated patients exhibited nausea compared with 0.3% of patients receiving placebo (P < 0.001). The incidence of treatment-emergent nausea during 6-month continuation of duloxetine treatment (80 mg/d, 2.1%; 120 mg/d, 1.3%) was similar to placebo (1.6%). Following abrupt discontinuation after 8 months of treatment, nausea was reported by 1.6% of patients receiving duloxetine 120 mg/d compared with 0% for those receiving duloxetine 80 mg/d and 0% for placebo. CONCLUSIONS: Duloxetine induced mild to moderate nausea in a subset of patients with MDD during treatment initiation. Nausea resolved rapidly with continued treatment. The incidence of duloxetine-induced nausea resembled that produced by paroxetine and fluoxetine.