Two amplification tests for the diagnosis of Chlamydia trachomatis infection, namely the ligase chain reaction (LCx) and the strand displacement assay (ProbeTec), were compared using samples from 1183 patients at sexually transmitted disease clinics. The overall prevalence of positive results was 8.0%, with agreement between the two assays of 98.8%. For endocervical, urethral and male urine samples, agreement was 99.3%, 99.4% and 97.7%, respectively. For ten discrepant samples, alternative amplification assays suggested that the LCx and ProbeTec assays gave erroneous results in six and four cases, respectively. Inhibition of amplification was observed with three (0.25%) urine specimens.