Complaints of premature ejaculation (PE) and its repercussions are culture-dependent. To report the measured intravaginal ejaculatory latency time (IELT) and the impact of PE in Kermanshah, Iran. From November 1996 through October 2008, 3,458 patients presented to us with self-diagnosed PE. In the first visit, after obtaining a psychosocial and sexual history, PE-specific bother was self-rated by the patients and the patients were advised to measure their IELTs over the next 2-3 weeks. In the second visit, the measured IELTs were reported by the patients. Patients' measured IELT and bother score. Age range was 17-80 years (mean 34.1, standard deviation [SD] 9.1, median 32). Sixty-five percent were married. Primary and secondary PE was reported by 2,105 (60.8%) and 1,353 (39.1%) patients, respectively. Occasional PE was reported by 36 (0.01%). Of those with multiple partners, 6% had partner-specific PE. IELT distribution was positively skewed. Anteportal ejaculation was reported by 97 (2.8%). In 3,458 self-reported PE patients, IELT was 1-15 seconds in 542 (15.7%), 16-30 seconds in 442 (12.8%), 31-60 seconds in 978 (28.3%), > 1 <or= 2 minutes in 551 (15.9%), > 2 <or= 5 minutes in 712 (20.6%), and >5 minutes in 136 (3.9%). IELTs of <or=1 minutes and <or=2 minutes were reported by 59.5% and 75.5%, respectively. Median bother score was 3 of 4. Pearson's correlation between IELT and bother (r = -0.607) was highly negative, with shorter IELTs being correlated with more bother. Six hundred forty-three patients (18.6%) always consumed opium to lengthen their IELTs. All 21 patients who started to use Tramadol as a PE treatment became addicted to it. Of 168 divorced couples due to PE, 23 divorced because the sexually dissatisfied wives became involved in extramarital affairs. Applying the Diagnostic and Statistical Manual of Mental Disorders criteria for PE and a cutoff IELT point of <or=2 minutes, and thus excluding the patients with IELTs of longer than 2 minutes, the patients with occasional PE, and the patients who reported no personal bother, of 3,458 self-reported PE patients, 2,571 (74.3%) had PE. Including the 97 patients with anteportal ejaculation, arithmetic mean IELT in 2,571 patients was 45.87 seconds, SD 36.1, median 40, and range 0-120. Excluding the anteportal ejaculation, arithmetic mean IELT in the remaining 2,474 patients was 47.67 seconds, SD 35.71, median 42, range 1-120, and geometric mean 31.06. PE has a devastating impact. In clinical practice, primary lifelong PE is by far the most common variety of PE. Occasional PE is a very rare cause of presentation. A significant proportion of the patients who consider themselves as premature ejaculators have IELTs of more than 2 minutes; the sexual medicine community should reach a consensus on the clinical diagnostic and therapeutic approach to these patients who consider themselves as premature ejaculators. The most important cause of bother in PE is the briefness of the ejaculatory latency, rather than the lack of control. Second-round ED (pseudoED) was reported for the first time. The worst type of PE, i.e., anteportal ejaculation, is not very rare; thus, geometric mean that excludes anteportal ejaculation is not suitable statistics to report the ejaculatory latency time; instead, median is the best measure of IELT reporting. Cutoffs of 0.5-2.5 percentiles are not suitable means to diagnose PE.