The authors analyzed the incidence of sexual dysfunction (SD) with different selective serotonin reuptake inhibitors (SSRIs; fluoxetine, fluvoxamine, paroxetine, and sertraline) and hence the qualitative and quantitative changes in SD throughout time in a prospective and multicenter study. Outpatients (192 women and 152 men; age = 39.6 +/- 11.4 years) under treatment with SSRIs were interviewed with an SD questionnaire designed for this purpose by the authors and that included questions about the following: decreased libido, delayed orgasm or anorgasmia, delayed ejaculation, inability to ejaculate, impotence, and general sexual satisfaction. Patients with the following criteria were included: normal sexual function before SSRI intake, exclusive treatment with SSRIs or treatment associated with benzodiazepines, previous heterosexual or self-erotic current sexual practices. Excluded were patients with previous sexual dysfunction, association of SSRIs with neuroleptics, recent hormone intake, and significant medical illnesses. There was a significant increase in the incidence of SD when physicians asked the patients direct questions (58%) versus when SD was spontaneously reported (14%). There were some significant differences among different SSRIs: paroxetine provoked more delay of orgasm or ejaculation and more impotence than fluvoxamine, fluoxetine and sertraline (chi 2, p < .05). Only 24.5% of the patients had a good tolerance of their sexual dysfunction. Twelve male patients who suffered from premature ejaculation before the treatment preferred to maintain delayed ejaculation, and their sexual satisfaction, and that of their partners, clearly improved. Sexual dysfunction was positively correlated with dose. Patients experienced substantial improvement in sexual function when the dose was diminished or the drug was withdrawn. Men showed more incidence of sexual dysfunction than women, but women's sexual dysfunction was more intense than men's. In only 5.8% of patients, the dysfunction disappeared completely within 6 months, but 81.4% showed no improvement at all by the end of this period. Twelve of 15 patients experienced total improvement when the treatment was changed to moclobemide (450-600 mg/day), and 3 of 5 patients improved when treatment was changed to amineptine (200 mg/day).