To evaluate the currently used Japanese Society for Medical Mycology (JSMM) method for testing the azole susceptibility of yeasts, the activities of fluconazole and itraconazole were tested against recently collected clinical isolates of Candida spp. (n=946) and compared with the National Committee for Clinical Laboratory Standards (NCCLS) M27-A2 microdilution reference method. Favorable correlation with the M27-A2 method was not seen for isolates of C. albicans, C. tropicalis or other Candida spp., particularly their trailing-growth isolates. However, the degree of correlation and agreement of MIC values were markedly improved when testing was performed by the modified JSMM method in which the end-point to be read was changed from IC80 (for the current JSMM method) to IC50. These results suggest that there is an urgent need to revise the current JSMM method.