To characterize 7-year changes in fasting serum insulin and glucose concentrations, the authors analyzed population-based data on 3,095 nondiabetic black and white men and women who were initially aged 18–30 years in the Coronary Artery Risk Development in Young Adults (CARDIA) Study. Measurements were made of fasting insulin (by an assay with little cross-reactivity to proinsulin) and fasting glucose on frozen samples from baseline and Year 7 examinations. Over the 7-year period, mean fasting insulin increased 10–25%, mean fasting glucose increased 7–10%, and mean body mass increased 7–12% across the four race-, sex-groups. The strongest predictor of both insulin increase and glucose increase was an increase in body mass over the 7 years. Adjusted for age and examination time period in race-, sex-specific repeated measures analyses, fasting insulin increased longitudinally by approximately 5 μ/mL per 5 kg/m2 increase in body mass index (p < 0.05). Adjusted for age and time period, fasting insulin increased over the 7 years by approximately 2.5 μU/mL per 0.08 unit increase in waist/hip ratio (p < 0.05), although this association was much stronger cross-sectionally. In a similar model, each 100 unit decrease in physical activity longitudinally predicted a 0.1–0.2 μU/ml increase in fasting insulin (p < 0.05 in black men only); this association was stronger and statistically significant in all race-, sex-groups cross-sectionally. Fasting insulin was not associated with energy intake either cross-sectionalty or longitudinally, but age- and time-adjusted associations of insulin change with change in several nutrients (e.g., energy from fat) were statistically significant in whites. The authors conclude that marked increase in weight in young adulthood adversely alters glucose and insulin metabolism, and that, if not reversed, this may lead to harmful health consequences in later life. Am J Epidemiol 1996; 144: 235–46.