Over 37 months, we conducted a prospective double-blind, randomized study in a cohort of 138 HIV-infected patients to compare the effect of two different strategies on the prevention and treatment of oropharyngeal candidiasis relapses and on the development of clinical and microbiological resistance to fluconazole. Each episode was treated with a 7 day course of fluconazole 200 mg/day, followed by secondary prophylaxis with fluconazole 150 mg once weekly matched to placebo. The duration of the double-blind phase of the study, from the day of randomization to the first primary end-point, was 347 +/- 186 days for the fluconazole group and 196 +/- 128 days for the placebo group (P < 0.001). A total of 33 patients remained relapse-free during the course of the study. Clinical failure was observed in a total of five patients (four in the fluconazole group, one in the placebo group; P = 0.15). Microbiological resistance was recorded in 12 patients (eight in the fluconazole group, four in the placebo group; P = 0.20). There were no significant treatment group differences in microbiological resistance whether comparisons were made for all cases or for cases up to 1 month post-study. In the few patients who developed clinical and/or microbiological resistance, the cumulative dose of fluconazole before entry into the study was a mean value of 8.6 g (compared with 2.9 g in patients without clinical and/or microbiological resistance). In summary, patients treated with secondary prophylaxis suffered fewer relapses of oropharyngeal candidiasis. Development of resistant candidiasis (clinical and/or microbiological) was rarely seen in either group and its incidence was not significantly different.