AIMS: To examine the detection limit of the ligase chain reaction kit for Chlamydia trachomatis, to study the inhibitory effect of phosphate on the ligase chain reaction, and to clarify the mechanism of inhibition. METHODS: Three reference serovars of C trachomatis--D/UW-3/Cx, F/UW-6/Cx, and L2/434/Bu--were used to test the sensitivity of the chlamydia ligase chain reaction. Comparison was made of the inhibition by phosphate before and after DNA amplification. Phosphate in up to 2.4 mM concentration was added to specimens of C trachomatis serovar D (1 to 50 inclusion forming units (IFU)/reaction) before DNA amplification to examine the concentration dependency of phosphate inhibition of the ligase chain reaction. RESULTS: The detection limits were 0.6 IFU/reaction for serovar D/UW-3/Cx and F/UW-6/Cx, and 0.4 IFU/reaction for L2/434/Bu. Phosphate inhibited the ligase chain reaction only when it was added before the amplification stage. The specimens containing chlamydia at 1 to 50 IFU/reaction were negative when the concentration of phosphate added at the prethermocycle stage was more than 1.2 mM. CONCLUSIONS: Ligase chain reaction analysis is a reliable method of diagnosing C trachomatis infection because of its high sensitivity. It would be clearly superior to the currently used methods if the problem of inhibitors could be eliminated. The mechanism of inhibition of the ligase chain reaction by phosphate was thought to be blockade of the amplification of the target DNA. The efficacy of the ligase chain reaction could be inhibited by phosphate in the urine, so duplicate dilution analysis of some negative specimens should be useful.