BackgroundDesvenlafaxine and duloxetine are selective serotonin and norepinephrine reuptake inhibitors. Their efficacy has not been directly compared using statistical hypotheses. This study evaluated the non-inferiority of desvenlafaxine extended-release (XL) to duloxetine in patients with major depressive disorder (MDD). MethodsIn this study, 420 adult patients with moderate-to-severe MDD were enrolled and randomly assigned (1:1) to receive 50 mg (once daily [QD]) of desvenlafaxine XL (n = 212) or 60 mg QD of duloxetine (n = 208). The primary endpoint was evaluated using a non-inferiority comparison based on the change from baseline to 8 weeks in the 17-item Hamilton Depression Rating Scale (HAMD17) total score. Secondary endpoints and safety were evaluated. ResultsLeast-squares mean change in HAM-D17 total score from baseline to 8 weeks was −15.3 (95% confidence interval [CI]: −17.73, −12.89) in the desvenlafaxine XL group and − 15.9 (95% CI, −18.44, −13.39) in the duloxetine group. The least-squares mean difference was 0.6 (95% CI: −0.48, 1.69), and the upper boundary of 95% CI was less than the non-inferiority margin (2.2). No significant between-treatment differences were found in most secondary efficacy endpoints. The incidence of the most common treatment-emergent adverse events (TEAEs) was lower for desvenlafaxine XL than for duloxetine for nausea (27.2% versus 48.8%) and dizziness (18.0% versus 28.8%). LimitationsA short-term non-inferiority study without a placebo arm. ConclusionsThis study demonstrated that desvenlafaxine XL 50 mg QD was non-inferior to duloxetine 60 mg QD in efficacy in patients with MDD. Desvenlafaxine had a lower incidence of TEAEs than duloxetine did.