The purpose of the study was to investigate whether the potency of effect on the beta cell differs with type of sulfonylurea (SU) and with degree of severity of diabetes. 12 maturity onset diabetics were classed according to fasting blood glucose (FBG) in three groups of 4 patients each. Each patient served as his own control. Glibenclamide, Gliquidone, Glusoxepide and placebo were administered in random order with degree dosage adjusted according to degree of severity of diabetes. All patients were given a standardized diet with 150 g carbohydrates per day. Fullday profiles of blood glucose, insulin, C-peptide and sulfonylurea level in serum were made on the third day under each preparation. Results showed that with proper nutrition and sufficient weight reduction, patients in group I (FBG 80--130 mg/dl) needed no oral medication and in fact showed a tendency towards hypoglycaemic episodes under oral therapy. In group II (FBG 130--200 mg/dl) the effect of nutrients on beta cell secretion appeared to be both enhanced and accelerated by SU administration. Satisfactory metabolic control was achieved with SU, but not with placebo. This group seems to represent the type of patient most likely to benefit from SU therapy. In spite of high dosage levels, satisfactory control was not achieved with SU in any patient in group III (FBG greater than 200 mg/dl). Depending on individual factors such as ketosis-proneness, vascular complications, age and psycho-social aspects, insulin administration should be considered for these patients. There were not differences between the individual SU preparations in the parameters studied. There was insufficient evidence for a pharmacokinetic differential diagnosis.