OBJECTIVE: To determine whether insulin resistance or insulin deficiency is primary in the pathogenesis of type II diabetes. DESIGN: Cohort analytic study of persons with normal glucose tolerance but with a high risk for developing type II diabetes (average follow-up time, 13 years). SETTING: Outpatients had an intravenous glucose tolerance test and were contacted periodically to ascertain diagnoses of diabetes. PARTICIPANTS: One hundred and fifty-five normal offspring, ranging in age from 16 to 60 years, of two parents with type II diabetes and 186 normal control subjects in the same age range who had no family history of diabetes. MEASUREMENTS AND MAIN RESULTS: Two phenotypic characteristics distinguished the offspring of diabetic parents from control subjects. They had slower glucose removal rates (Kg) (P less than 0.01) and higher insulin levels (fasting and during the second phase of insulin response to intravenous glucose; P less than 0.0001) than did control subjects, even after adjustment for differences in obesity. Sixteen percent of the offspring developed type II diabetes. Mean Kg at baseline was 1.7%/min among offspring who subsequently developed diabetes, 2.2%/min among offspring who remained nondiabetic, and 2.3%/min among control subjects. Corresponding means for first-phase insulin were 498, 354, and 373 pM, respectively, whereas second-phase insulin means were 329, 117, and 87 pM, respectively. In multivariate analysis, low Kg and high serum insulin levels independently increased the risk for developing diabetes among the offspring of diabetic parents. CONCLUSIONS: One to two decades before type II diabetes is diagnosed, reduced glucose clearance is already present. This reduced clearance is accompanied by compensatory hyperinsulinemia, not hypoinsulinemia, suggesting that the primary defect is in peripheral tissue response to insulin and glucose, not in the pancreatic beta cell.