Depression is a common cause of sexual dysfunction, but also antidepressant medication is often associated with sexual side effects. This article includes two related studies. The first double-blind, placebo-controlled study was conducted in men with lifelong rapid ejaculation and aimed to assess putative differences between the major selective serotonin reuptake inhibitors (SSRIs) (fluoxetine, fluvoxamine, paroxetine, and sertraline) with regard to their ejaculation-delaying effect. Sixty men with an intravaginal ejaculation latency time (IELT) of 1 minute or less were randomly assigned to receive fluoxetine 20 mg/day, fluvoxamine 100 mg/day, paroxetine 20 mg/day, sertraline 50 mg/day, or placebo for 6 weeks. During the 1-month baseline and 6-week treatment periods, the men measured their IELT at home using a stopwatch. The trial was completed by 51 men. During the 6-week treatment period, the geometric mean IELT in the placebo group was constant at approximately 20 seconds. Analysis of variance revealed a between-groups difference in the evolution of IELT delay (p = 0.0004); in the paroxetine, fluoxetine, and sertraline groups there was a gradual increase to about 110 seconds, whereas in the fluvoxamine group, IELT was increased to only approximately 40 seconds. The paroxetine, fluoxetine, and sertraline groups differed significantly (p < 0.001, p < 0.001, p = 0.017, respectively) from placebo but the fluvoxamine group did not (p = 0.38). Compared with baseline, paroxetine exerted the strongest delay in ejaculation, followed by fluoxetine and sertraline. There was no clinically relevant delay in ejaculation with fluvoxamine. In men with lifelong rapid ejaculation, paroxetine delayed ejaculation most strongly, whereas fluvoxamine delayed ejaculation the least. The second double-blind, placebo-controlled study was carried out in men with lifelong rapid ejaculation (IELT < or = 1 minute) and in men with lifelong less-rapid ejaculation (IELT > 1 minute) to investigate whether data about SSRI-induced delayed ejaculation in men with rapid ejaculation may be extrapolated to men with less-rapid ejaculation. After measurement of IELT at home (using a stopwatch) during a 1-month baseline assessment, 32 men with an IELT of 1 minute or less (group 1) or more than 1 minute (group 2) were randomly assigned to receive paroxetine 20 mg/day or placebo for 6 weeks in a double-blind manner. Patients continued to measure their IELTs at home during the 6 weeks of the study. At baseline, 24 patients consistently had IELTs of one minute or less (group 1), and eight patients had IELTs of more than 1 minute (group 2). The geometric mean IELT was 14 seconds in group 1 and 83 seconds in group 2. Twelve patients in group 1 and five in group 2 were randomized to the paroxetine 20 mg/day. The percentage increase in the geometric mean IELT compared with baseline in patients treated with paroxetine was 420% (95% confidence interval [CI], 216-758%) in group 1 and 480% (95% CI, 177-1,118%) in group 2 (p = 0.81). After 6 weeks of treatment with paroxetine, the geometric mean IELT was 92 seconds in group 1 and 602 seconds in group 2 (p < 0.001). Therefore, the paroxetine-induced percentage increase in IELT seems to be independent of the baseline IELT. This suggests that ejaculation-delaying side effects of some SSRIs investigated in men with lifelong rapid ejaculation may be generalized to men with less-rapid ejaculation.