Two commercial antifungal susceptibility testing systems (Fungitest and Neo-Sensitabs) were compared with the M27T-NCCLS reference broth microdilution method using one hundred isolates of Candida sp. and Crptococcus neoformans. Six different antifungal drugs were tested: amphotericin B, 5-fluorocytosine, fluconazole, itraconazole, ketoconazole and miconazole. The overall agreement between the Fungitest and the reference methods was much better than between the Neo-Sensitabs and the reference methods: the agreement for the Fungitest ranged from 100% for amphotericin B to 76.7% for itraconazole whereas for the Neo-Sensitabs, it ranged from 90.4% for amphotericin B to 36% for ketoconazole. For the total number of tests performed with Neo-Sensitabs, there were 37.8% of discrepancies with the reference method whereas for the tests performed with Fungitest, there was only 16.5% of discrepancies. Major discrepancies, defined as results that classified an isolate as susceptible by one method and resistant by another, occurred in 21 cases for the Neo-Sensitabs test and only in four cases with the Fungitest, namely 0.6% of the cases. We conclude that the Fungitest method constitutes a simple and reliable procedure for antifungal drug susceptibility testing