This study tests the possibility to avoid the use of phenformin in 40 type 2 (non insulin dependent) diabetic patients treated with the commercial sulphonylurea-phenformin combinations. In diabetic patients treated with sulphonylureas and phenformin at low dosage (glibenclamide 5 mg and phenformin 50 mg) it was possible to maintain good glycometabolic control using only the sulphonylurea gliclazide (160 mg/die). The diabetic patients on treatment with sulphonylureas and phenformin at higher dosage (glibenclamide 7.5 mg and phenformin 75 mg) may further improve their metabolic control when transferred to a gliclazide-benfluorex combination 160 mg and 300 mg/die, respectively. These results suggest the possibility of withdrawing or replacing phenformin in the therapy of type 2 diabetic patients without modifying their glycometabolic control.