This is a retrospective study of clinical experience collected at the University Clinical Hospital over a 19-year period. Semen samples were analyzed according to WHO criteria. In the postmasturbatory urine, sperm count was performed. Data were expressed as total sperm number in urine (TSNU) and using a retroejaculation index. Patients were categorized into four groups according to the presence of sperm in the studied samples: a) in semen and urine; b) only in urine; c) only in semen; d) neither in semen nor in urine. A control group included nonretroejaculator patients. Retroejaculator patients are those whose TSNU is superior to 3.8 x 10(6) and the RI superior to 2.16%. While diagnosing retroejaculation, the only presence of sperm in the postmasturbatory urine is not adequate. The proposed index added to total sperm number in urine and semen volume may identify true retroejaculator patients.