Fluconazole (UK 49,858), a new orally administered bis-triazole, was compared with ketoconazole for activity in synthetic broth dilution susceptibility tests against Candida albicans and also in treatment of experimental systemic candidal infections in rats. In vitro studies indicated that fluconazole activity is less sensitive to acidic medium than is that of ketoconazole. At physiologic pH, fluconazole was approximately 16-fold less active than ketoconazole against 35 representative isolates of C. albicans. Two additional isolates (K-1 and K-3) recovered from patients who had failed ketoconazole therapy were 32- to 64-fold more resistant than the median of each drug for other isolates. In animal studies, fluconazole was very effective in prolonging survival of rats infected with a representative candidal strain. With an inoculum sufficient to kill 29 of 38 sham-treated animals, only 1 of 18 animals treated with 0.5 mg of fluconazole per kg per day died compared with 13 of 20 animals treated with 10.0 mg of ketoconazole per kg per day. However, when similar fluconazole treatment was administered to rats infected with the more resistant strain, K-1, no prolongation of survival was found. Thus, in vivo and in vitro results between strains correlated well for fluconazole. However, in comparing results between drugs, ketoconazole was 16-fold more active in vitro and fluconazole was 20-fold more active in vivo. This discrepancy may be due to drug distribution, modes of drug metabolism, or other pharmacologic differences between the two agents.