The effects of prolonged itraconazole exposure on the susceptibility of Candida albicans isolates to itraconazole and fluconazole have not been well characterized. A recent placebo-controlled study of long-term itraconazole antifungal prophylaxis in persons with advanced human immunodeficiency virus infection afforded the opportunity to address this question. Mucosal Candida sp. isolates were obtained from subjects who developed oropharyngeal or esophageal candidiasis, and in vitro susceptibilities of the last isolate obtained at removal from the study as a prophylaxis failure were compared in itraconazole and placebo recipients. More subjects in the placebo group (74 of 146 [51%]) than in the itraconazole group (51 of 149 [34%]) developed mucosal candidiasis (P = 0.004). A total of 112 isolates were recovered from 56 of the 74 (76%) subjects with mucosal candidiasis assigned to the placebo group, compared to 97 isolates from 45 of the 51 (88%) subjects in the itraconazole group. C. albicans accounted for 98% of isolates in the placebo group and 89% of isolates in the itraconazole group. The itraconazole MIC at which 50% of the isolates tested were inhibited (MIC(50)) for last-episode isolates from the itraconazole group was 0.125 microg/ml compared to 0.015 microg/ml for the placebo group subjects, P = 0.0001. The MIC(50) of fluconazole for the last isolates from the itraconazole group was 1.5 microg/ml compared to 0.5 microg/ml for the placebo subjects (P = 0.005). A lower proportion of isolates recovered from subjects on itraconazole therapy were classified as susceptible to itraconazole (63%) compared to isolates from the placebo group (96%) (P = 0.001). Similarly, a lower proportion of C. albicans isolates from subjects on itraconazole therapy were susceptible to fluconazole (78%) compared to isolates from the placebo group (96%) (P = 0.01). Also, the proportion of isolates that were not fully susceptible to itraconazole or fluconazole was greater in patients assigned to the itraconazole group than the placebo group (itraconazole susceptibility, 37 and 4%, respectively (P = 0.001); fluconazole susceptibility, 23 and 4%, respectively (P = 0.01). In conclusion, long-term itraconazole prophylaxis in patients with AIDS is associated with reduction in susceptibility to itraconazole and cross-resistance to fluconazole.