The primary objective of this post-hoc analysis was to evaluate the effect of short-term treatment with desvenlafaxine versus placebo on sexual dysfunction (SD), assessed from Arizona Sexual Experiences Scale scores, in adult outpatients with major depressive disorder. Data from three randomized, double-blind, placebo-controlled trials of 50 or 100 mg/day desvenlafaxine for major depressive disorder were pooled. SD status, determined from Arizona Sexual Experiences Scale scores, was assessed at baseline and week 8, last observation carried forward. Subgroup analyses addressed the effects of sex, baseline SD, and antidepressant response. At week 8, last observation carried forward (n=1562), SD rates were 54, 47, and 49% for 50 mg/day desvenlafaxine, 100 mg/day desvenlafaxine, and placebo, respectively [adjusted odds ratios (95% confidence interval) vs. placebo: 1.205 (0.928, 1.564) and 1.129 (0.795, 1.604), respectively]. The treatment by baseline SD interaction approached statistical significance (P=0.0663), mainly driven by poorer scores for desvenlafaxine versus placebo in the 100 mg group. Treatment by sex interactions were not statistically significant. Small but statistically significant treatment by sex interactions were observed for sex drive (P=0.0011) and ease of erection/lubrication (P=0.0151). Although there was no overall effect of desvenlafaxine on SD, a treatment by baseline SD interaction was suggested for 100 mg desvenlafaxine.