Diabetes and obesity are epidemic in the Pima Indians of the Southwestern United States, and the prevalence of diabetes is increasing. The most likely link between obesity and diabetes is tissue insulin resistance. If obesity is defined as an excess of body fat, then it can only be accurately assessed by measurements of body composition and not by approximations such as body mass index or percent of ideal weight. To compare the metabolic data of individuals of varying size, an accurate measure of metabolic size is needed. Total body weight is not an appropriate means of comparing individuals since obese subjects have a greater proportion of nonmetabolizing mass (triglyceride). Body surface area shows a sex difference, and this may distort data if both sexes are present. From studies of metabolic rate we have determined that metabolic rate is indirectly proportional to the fat-free mass plus 18 kg, and we suggest that this weight can be equated with metabolic size. Glucose storage in skeletal muscle appears to be important in the disposal of an intravenous glucose load. Consistent with its role in glycogen storage, glycogen synthase enzyme is activated in proportion to the ability to dispose of glucose during a hyperinsulinemic, euglycemic clamp. The role of glycogen synthase is most notable at supraphysiological plasma insulin concentrations; and since glucose uptake at these insulin concentrations is highly familial independent of the degree of obesity, we suggest that there may be a specific genetic defect expressed in skeletal muscle that reduces insulin responsiveness in some subjects. The lack of correlation between 24 hour respiratory quotient measured in a metabolic chamber (a measure of the proportion of fat derived calories) and degree of obesity indicates that in obese Pima Indians insulin resistance is not due to an inhibition of glucose metabolism by free fatty acids (glucose-fatty acid-ketone cycle). Obesity is associated with an increase in fat-free mass almost kilogram- for kilogram with fat mass when compared to the lean state. A role for this increase in fat-free tissue in producing insulin resistance has been given insufficient attention in the past. With an increase in fat-free mass, muscle cells are hypertrophied and capillaries in muscle are more widely spaced. We propose that these biophysical changes in muscle mediate, at least in part, the effects of obesity to produce a reduction in insulin sensitivity and the abnormal kinetics of insulin action found in the obese. We suggest therefore that insulin resistance is a combination of a genetic defect and obesity-induced changes in the biophysical properties of skeletal muscle. These defects may in turn lead to the development of non-insulin-dependent diabetes mellitus.