Eight patients aged 57-86 years with non-insulin-dependent diabetes mellitus (NIDDM) taking long-term glibenclamide treatment (5-15 mg/day) were given oral trimethoprim-sulphamethoxazole due to an acute bacterial infection. After 4-6 days of combined treatment, the total plasma glibenclamide concentrations were determined every second hour for 12 h, and the area under the curve (AUC) was computed. A second plasma glibenclamide profile was obtained 2-4 weeks after withdrawal of trimethoprim-sulphamethoxazole. The levels of plasma glibenclamide did not differ between the two occasions, nor did the simultaneously determined levels of blood glucose and plasma insulin. It is concluded that there is no consistent pharmacokinetic interaction between glibenclamide and trimethoprim-sulphamethoxazole in NIDDM patients.