The combination of a sulfonylurea with a biguanide improves the pancreatic beta-cell insulin secretion and the insulin utilization in peripheral tissues in NIDDM. This open, crossover, randomised and prospective study was designed to compare the effects of the fixed combination glibenclamide-metformin (GL-METF)-2.5 and 400 mg respectively, with the fixed combination glibenclamide-phenformin (GL-PHEN)-2.5 and 25 mg respectively, on NIDDM diabetes control. Thirty NIDDM patients, in ideal metabolic control, who were being treated with GL-PHEN were divided in two groups. One group received GL-PHEN for 12 weeks followed by 12 weeks treatment with GL-METF and the reverse treatment was given to the second group. A statistically significant decrease of post-prandial blood glucose (p = 0.034) and glycosylated haemo-globin (p < 0.02) values was observed under GL-METF treatment compared to those with GL-PHEN. The values of lactic acid were within normal limits during both treatments. The insulin secretion after breakfast was similar with both drug compounds. The BMI of the patients remained the same during a follow-up study of 24 weeks. Lipid metabolism did not change significantly during the trial and the safety parameters (renal and liver function, full blood count) remained unchanged. In conclusion, the administration of GL-METF leads to better diabetes control in NIDDM patients compared to that of GL-PHEN.