Patients with premature ejaculation (PE) were subdivided into primary (PPE), individuals who had suffered from PE since the beginning of their sexual lives, and secondary (SPE), those who developed the condition after years of satisfactory sexual functioning. PPEs differed from SPEs on a number of clinical and psychometric variables. Clinically SPEs were significantly more likely to manifest a coexisting erectile disorder, reduction in sex drive, and a decrease in arousal during sexual stimulation than SPEs. They were significantly less likely to report high levels of anxiety during coitus. Psychometrically, on the Derogatis Sexual Functioning Inventory, PPEs were significantly more "impaired" than the SPEs as reflected by scores on the Symptoms and Satisfaction scales and the GSSI. They were significantly less impaired on measures of sex drive (e.g., Drive and Fantasy). On the Hamilton Anxiety Rating Scale (HRAS), the PPEs scored as significantly more anxious than the SPEs. The findings suggest that dichotomizing PE into PPE and SPE may be clinically useful, and may have etiologic treatment and prognostic implications.