Although generally accepted standards exist for routine semen analysis, recent methods of assessing reactive oxygen species (ROS) in human semen lack a standardized protocol. The purpose of this study was to investigate (1) the relationship between ROS level and the time interval between semen collection and analysis, and (2) the effect of sperm concentration on the level of ROS formation. Semen specimens from men (n = 40) consulting for infertility treatment were divided in two groups: in 20, routine semen analysis was performed and ROS formation evaluated at 1, 3, 5, and 24 h after semen collection; in the other 20, ROS formation was evaluated at four sperm concentrations (60, 30, 15, and 7.5 x 10(6)/mL). White blood cell (WBC) concentration was assessed before ROS measurement using a myeloperoxidase staining technique (Endtz test). ROS level was measured by a chemiluminescence method. ROS formation decreased significantly over time. The mean ROS level 343.4 (1 h), 133.5 x 10(4) cpm (3 h, p = .004), 66.0 x 10(4) cpm (5 h, p < or = .001), and 22.2 x 10(4) cpm (24 h, p < or = .001), respectively. In the first group of 20 specimens, 14 were positive for ROS formation at 1 h after collection, and 4 of these were positive for the Endtz test (> 1 x 10(6) WBC/mL). The number of ROS-positive specimens after 3, 5, and 24 h was eight, six, and two, respectively. In the second group, eight patients were positive for ROS formation at 1 h after collection.(ABSTRACT TRUNCATED AT 250 WORDS)