An attempt was made to use uncentrifuged segmented urine specimens and Kova-Slide II with grid as a quantitative technique for the diagnosis of nongonococcal urethritis. Of the 100 men with urethral symptoms, 54 had fewer than five polymorphonuclear leukocytes per high-power field (x1000) in their gram-stained urethral smear. On the basis of examination of segmented urine and Kova-Slide test, 41 of these 54 patients were diagnosed as having nongonococcal urethritis. None of the 13 patients considered to be free of urethritis on the basis of segmented urine tests had a Chlamydia trachomatis--positive culture, and all became asymptomatic without treatment. All 46 patients who had more than five polymorphonuclear leukocytes per high-power field (x1000) also had positive segmented urine tests. The rate of isolation of C. trachomatis was similar for both groups of tests. This study revealed that patients with lower polymorphonuclear leukocyte counts in the first 10 ml of urine passed (voided-bladder urine 1) also had cultures positive for C. trachomatis. None of the 50 patients in the control group had polymorphonuclear leukocytes in either voided-bladder urine 1 or in the midstream specimen (voided-bladder urine 2). All controls had cultures negative for C. trachomatis.