In order to verify whether fluconazole has a prophylactive effect against the occurrence of cryptococcosis in HIV-infected patients and to identify other factors capable of increasing or reducing the risk of this infection, we arranged a case-control study of 17 patients with cryptococcal infection. 34 controls were selected, matched by presence of an AIDS-defining event, CD4 cell count, and date of T-cell phenotyping. No significant difference in exposure to fluconazole, in total days of treatment, or in total dose administered was observed between cases and controls. However, control patients took a significantly higher average daily dosage of fluconazole and a linear tendency in risk reduction (p = 0.04) in relation to increasing dosage was observed. Antiretroviral therapy and an average daily fluconazole dose exceeding 150 mg both each reduced the risk of a cryptococcal infection.