During a randomized double-blind placebo-controlled study testing the efficacy of itraconazole for prophylaxis of systemic and mucosal fungal infections in patients with acquired immune deficiency syndrome, 298 patients were enrolled with 295 evaluable. Of those, 46 patients were considered prophylaxis failures because of recurrent oral or esophageal candidiasis. Oropharyngeal fungal cultures were taken at the time of suspected thrush or Candida esophagitis, but not at baseline. All of the Candida spp. isolates were cultured on CHROMagar Candida medium then identified using API 20 AUX strips. Antifungal susceptibility testing was performed following the National Committee for Clinical Laboratory Standards M-27A guidelines. Sequential isolates were genotyped using randomly amplified polymorphic DNA. Polymerase chain reaction fingerprints were generated using two repetitive sequence primers, (GGA)7 and (GACA)4. The study group consisted of 23 patients, nine from the itraconazole arm and 14 from the placebo arm, who were prophylaxis failures and had more than two C. albicans isolates. Five of 23 had isolates showing a > or =4-fold reduction in susceptibility; four of these patients were in the itraconazole prophylaxis arm and one was in the placebo arm. Three of the five had yeast isolations showing changes in banding patterns over time. Such changes may indicate genetic changes in the same strain that could be linked to acquired resistance to itraconazole, or acquisition of a new strain, or emergence of a previously minor component of the original population.