OBJECTIVE: TO determine whether salivary concentrations of ketoconazole and fluconazole may explain the apparent disparity between in vitro activity and clinical efficacy observed with these drugs. DESIGN: Healthy subjects received a single oral dose of ketoconazole 400 mg or fluconazole 100 mg in a randomized, crossover fashion. Saliva was collected at O, 1, 2, 3, 6, 12, and 24 hours. Blood samples were obtained at 2 and 24 hours. Salivary concentrations and plasma concentrations for each drug were determined by HPLC. Minimum inhibitory concentration (MIC) testing was determined in triplicate on 6 clinical isolates of Candida albicans, and times over the median MIC values were calculated. PARTICIPANTS: Eight subjects completed the study. RESULTS: The mean (+/- SD) peak salivary concentration for ketoconazole was 0.119 +/- 0.050 mu g/mL at 3 hours; no subject had a detectable ketoconazole salivary concentration at 24 hours. At 2 and 24 hours, mean ketoconazole plasma concentrations were 7.64 +/- 3.87 and 0.11 +/- 0.05 mu g/mL, respectively. The saliva to plasma concentration ratio at 2 hours was 0.01. The mean peak salivary concentration of fluconazole was 2.56 +/- 0.34 mu g/mL at 3 hours. At 24 hours, the mean salivary concentration was 1.44 +/- 0.33 mu g/mL. At 2 and 24 hours, mean fluconazole plasma concentrations were 4.39 +/- 3.33 and 3.72 +/- 2.83 mu g/mL, respectively. The saliva to plasma concentration ratio at 2 hours was 0.55. Median MIC values were 0.0625 mu g/mL, (range 0.0313-0.125) for ketoconazole and 0.25 mu g/mL (range 0.125-0.5) for fluconazole. Calculated times over which ketoconazole and fluconazole exceeded the median MICs in saliva were approximately 13 and greater than 24 hours, respectively. CONCLUSIONS: After a single oral dose, fluconazole achieved higher salivary concentrations than did ketoconazole. This may explain the increased clinical efficacy of fluconazole in the treatment of oropharyngeal-esophageal candidiasis when compared with ketoconazole.