The search for the genetic basis of NIDDM has magnified the need for an efficient representation of the pre-NIDDM phenotype. The overall goal is to relate specific mutations on the genome to specific changes in physiologic function which lead to NIDDM. Unfortunately, there is still not a clear understanding of the molecular cause of NIDDM in most individuals. Therefore, one must take an alternative approach: to express in quantitative terms the various tissue processes which determine the ability to regulate the blood glucose in fasting and after carbohydrate administration. A minimal list of such processes includes the provision of glucose by the liver, insulin sensitivity, insulin secretion, and glucose effectiveness. The latter function is the ability of glucose per se to enhance glucose disappearance from blood, independent of a dynamic insulin response. Approaches to measuring the list of functions which determine the glucose tolerance are reviewed: they include the minimal model method, which quantitates insulin sensitivity (Sl) and glucose effectiveness (SG), and a combined model approach, which measures insulin secretion. These methods are being developed for large populations. Such a development is important for elucidating the causes of reduced glucose tolerance in populations, and examining the relation between such causes and outcomes including diabetes and cardiovascular disease. Of particular importance for diabetes development is the characteristic hyperbolic relationship between insulin secretion and insulin action. This relationship, the "hyperbolic law of glucose tolerance' indicates that insulin secretion can only be assessed in terms of the ambient degree of insulin sensitivity. By applying this principle, it is clear that latent pancreatic islet-cell dysfunction has been underestimated, and may be significant even in subjects with impaired glucose tolerance. Finally, new explorations of insulin control of liver glucose output indicate that this process may be under the control of free fatty acids. The latter realization indicates that the insulin effect on lipolysis is what is critical for determination of glucose output in the fasting state, and that insulin resistance at the level of the adipocyte may determine the extent of fasting hyperglycaemia, and may be an important factor in the overall phenotype in prediabetic and NIDDM individuals.