Sulphonylureas have remained the mainstay of oral therapy for type 2 (non-insulin-dependent) diabetes mellitus (NIDDM). They stimulate insulin release from pancreatic beta cells. Pharmacokinetic differences between the various sulphonylureas are of clinical importance in terms of the time to onset of action, timing of drug administration in relation to food intake, magnitude and duration of the glucose-lowering effect and the risk of serious hypoglycaemia. Recent studies with improved analytical sensitivity have shown that the elimination half-life of glibenclamide is longer than previously thought and that 2 metabolites of glibenclamide have significant hypoglycaemic activity. Furthermore, single dose studies in healthy volunteers using an integrated pharmacokinetic-pharmacodynamic model have identified clear concentration-effect relationships for both glibenclamide and its metabolites after oral and intravenous administration. Under multiple dose conditions, kinetic-dynamic relations have been identified with shorter-acting drugs in dosages that give discontinuous sulphonylurea exposure. However, at continuous exposure, i.e. sustained 24-hour therapeutic concentrations in plasma, there is evidence indicating the development of tolerance, which may be caused by downregulation of beta cell sensitivity. As more sophisticated concentration-effect studies appear, it has become evident that currently recommended maximum daily doses of many sulphonylureas are too high.