Like many other hormones insulin is released in a pulsatile manner, which results in oscillatory concentrations in peripheral blood. The oscillatory pattern is believed to improve release control and to enhance the hormonal action. Insulin oscillates with a slow ultradian periodicity (approximately 140 min) and a high-frequency periodicity (approximately 6-10 min). Only the latter will be discussed in this review, which focusses almost exclusively on human data. Probably at least 75% of insulin secretion is released in a very regular pulsatile fashion in healthy people. In contrast, type 2 diabetic subjects exhibit in general irregular oscillations of basal plasma insulin. Furthermore, disturbed pulsatile insulin release is also a common feature in people prone to develop diabetes e.g. first-degree relatives of patients with type 2 diabetes. Tiny glucose oscillations (approximately 0.3mM) are capable of easily governing or entraining insulin oscillations in healthy people. This differs from type 2 diabetic individuals underlining a profound disruption of the beta-cells in type 2 diabetes to sense or respond to physiological glucose excursions. A pivotal question is how approximately 1,000,000 islets each containing from a few to several thousand beta-cells can be coordinated to secrete insulin in a pulsatile manner. Coordination is extremely complex involving the intrapancreatic neural network, the intraislet communication and metabolic oscillations in the individual beta-cell itself, but our understanding is still rather rudimentary. It is important to realize how antidiabetic treatment influences high-frequency insulin pulsatility. Only a few studies have been performed, but very recently it has been demonstrated that acute as well as long-term administration of the sulfonylurea compound gliclazide results in a substantial amplification (approximately 50%) of the pulsatile insulin secretion in type 2 diabetes. The fraction between pulsatile vs nonpulsatile insulin secretion is stable, which may be essential for controlling glucose and lipid homeostasis in type 2 diabetes. The influence of repaglinide, thiazolidinediones and a potential future antidiabetic compound (GLP-1) on pulsatile insulin secretion is also discussed briefly. Evaluation of high-frequency insulin pulsatility may be an important player in future tailoring of antidiabetic drugs and may turn out to be relevant as a predictor of type 2 diabetes in people at high risk for developing the disease.