1. Cytochrome P450-mediated bioactivation of sulphamethoxazole to a hydroxylamine has been implicated in the hypersensitivity reactions associated with co-trimoxazole administration. Inhibiting the formation of the hydroxylamine may be one method of preventing the high frequency of toxicity which is observed in HIV-infected patients. Therefore, in this study, we have investigated the ability of fluconazole and ketoconazole, known cytochrome P450 inhibitors, to inhibit the formation of sulphamethoxazole hydroxylamine. 2. Ten healthy male volunteers were given co-trimoxazole (800 mg sulphamethoxazole and 160 mg trimethoprim) alone or 1 h after either fluconazole (150 mg) or ketoconazole (200 mg) in a randomized fashion with a washout period of at least 1 week between each phase. Urine was collected for 24 h, and sulphamethoxazole and its metabolites were quantified by electrospray LC-MS. 3. Ketoconazole had no effect on the urinary recovery of sulphamethoxazole or any of its metabolites. In contrast, fluconazole significantly (P < 0.001) inhibited the formation of sulphamethoxazole hydroxylamine by 50.0 +/- 15.1%. Fluconazole also inhibited the oxidation of sulphamethoxazole to the 5-methylhydroxy and 5-methylhydroxy acetate metabolites by 69.9 +/- 15.8% and 64.0 +/- 12.0%, respectively, but had no effect on the amount of sulphamethoxazole, N4-acetyl sulphamethoxazole, or sulphamethoxazole N1-glucuronide excreted in urine. 4. The potential clinical benefit of using fluconazole to prevent hypersensitivity to co-trimoxazole in patients with AIDS needs to be assessed in a prospective study using both metabolite formation and the clinical occurrence of adverse reactions as end-points.