50 type II diabetics with secondary failure of a sulfonylurea (SU) treatment have been included in a randomised controlled trial by assigning them either to a Metformin + SU or insulin + SU treatment. At study entry the metabolic control of patients had to be insufficient, as demonstrated by increased fasting or postprandial plasma glucose levels (greater than 130 resp. greater than 180 mg/100 ml). The study lasted for a period of 12 months in which a Metformin treatment of 850, 1,700 or 2,550 mg/day or an insulin therapy of 12-40 IU was added to the existing maximal dose of SU. Both the Metformin and the insulin dosage were adapted to the response of metabolic control. Parameters of metabolic control and security were monitored regularly every two months. 16 patients in the Metformin + SU-group and 19 patients in the insulin + SU-group were observed for the total study period of 12 months. In both groups there was a similar development of metabolic parameters. Stimulated C-peptide reduced in both groups indicating an increase of peripheral insulin sensitivity under both treatment regimes.