Our objective was to compare 3 deoxyribonucleic acid (DNA) amplification methods for the diagnosis of chlamydial infection with an enhanced enzyme immunoassay (EIA) method for antigen detection in urine samples, from men with non-gonococcal urethritis (NGU) attending a busy inner city genitourinary medicine centre. Urethral swabs and urine samples were collected from 346 male patients with NGU attending the clinic. All swabs and urines were tested for chlamydial infection (CT) using the EIA (Dako PCE immunoassay). Three aliquots of the urine samples were stored immediately at -70 degrees C for subsequent testing by: Amplicor polymerase chain reaction (PCR) (Hoffmann-La Roche, Switzerland); the amplified Chlamydia trachomatis assay (AMP CT) using transcription mediated amplification (TMA) (GenProbe, USA); and BDProbeTecET using the strand displacement assay (SDA) (Becton Dickinson, USA). The positive rate for the 3 amplified assays PCR, TMA and SDA (on urine) was 88/346 (25.4%), 80/346 (23.1%) and 88/346 (25.4%), respectively compared to 56/346 (16.2%) by EIA on urethral swabs, the current means of diagnosis in this laboratory. Thirty-one samples were positive in 2 or more of the amplification assays but negative in the EIA, 50 positives (53% sensitivity) detected in the urine samples by the EIA assay were detected by all 3 of the amplified assays. Three samples were positive by PCR only, 5 were positive by TMA only and 7 were positive by SDA only. DNA amplification assays are superior to standard immunoassays for the diagnosis of C. trachomatis infections in urine samples. Urine samples are suitable for use in these amplified assays to detect C. trachomatis. Freezing of samples before testing reduces the rate of inhibition reported in other published studies.