The proposed standard of the Antifungal Susceptibility Testing Subcommittee of the European Committee on Antibiotic Susceptibility Testing (AFST-EUCAST) and the E-test procedures were compared with the NCCLS reference broth microdilution method for voriconazole and caspofungin susceptibility testing of 102 clinical Candida species and Saccharomyces cerevisiae isolates. The voriconazole MIC at which 50% of strains were inhibited (MIC(50)) was < or =0.125 mg/liter for all yeast species except for Candida glabrata and Candida krusei, which yielded MIC(50) values of 0.25 to 1 mg/liter depending on the method. Caspofungin exhibited in vitro activity (MIC(50) of < or =0.125 to 2 mg/liter) against all yeast species except for Candida guilliermondii. The agreements between MICs within +/-2 dilutions obtained by the NCCLS method and the EUCAST standard were 97% for voriconazole and 96% for caspofungin. Intraclass correlation coefficients were statistically significant (P < 0.05). The agreements between voriconazole MICs provided by the E-test and the NCCLS and between the E-test and the AFST-EUCAST method were 100 and 90%, respectively. Because of lower caspofungin MICs provided by the E-test, the agreement was slightly poorer with the NCCLS method (89%) than with the AFST-EUCAST procedure (94%). Both the EUCAST and the E-test procedures can be reliable techniques for susceptibility testing of yeasts to voriconazole and caspofungin.