Considerations of the pathophysiology of non-insulin dependent diabetes mellitus (NIDDM) usually focus on the respective roles of the so-called triumvirate-beta cell, muscle and liver [1]. Often overlooked in this context is the role of the adipose tissue, and attention is usually addressed to consideration of studies in which isolated adipocytes were used as a surrogate for muscle in studies of insulin action. The goal of this presentation will be to develop a radically different hypothesis, and marshal evidence that it is the loss of normal regulation of adipose tissue that plays the central role in both the hyperglycaemia and the dyslipidaemia that characterizes patients with NIDDM.