The use of nucleic acid amplification tests for the diagnosis of C. trachomatis has made it possible to send urine samples instead of urethral swab specimens to the laboratory. The sensitivity is very high, but not 100%, and we continue to perform a test for urethritis at our STD clinic. The aim of this study was to compare the performance of two alternative tests in the diagnosis of urethritis as predictors of C. trachomatis infection: the leukocyte esterase (LE) dipstick test of first-voided urine and polymorphonuclear leukocyte counts in a methylene blue-stained (MBS) urethral smear. Urine samples from 480 male patients attending an STD clinic were analysed using the LE test and LCR assay for C. trachomatis; urethral samples were analysed with MBS urethral smear and LCR. The majority (75.8%) of the 480 patients examined were asymptomatic. Chlamydial infection was detected in 50 patients. The sensitivity, specificity and positive predictive value of the LE test for predicting C. trachomatis infection were 46.0, 91.6 and 39.0%, respectively, among all patients examined and 25.9, 95.8 and 33.3%, respectively, among the asymptomatic patients. The corresponding values for the MBS urethral smear were 76.0, 82.1 and 33.0% among all patients and 63.0, 89.6 and 32.7% among the asymptomatic patients. At our STD clinic we chose to perform the examination of MBS urethral smears in the diagnosis of urethritis because of its higher sensitivity relative to the LE test for predicting C. trachomatis.