OBJECTIVE: To determine the relation of weight gain in middle age with diabetes-related medical utilization and costs in older age. METHODS: Study cohort included 1,063 men from the Chicago Western Electric Study: ages 40-55, not underweight (body mass index <18.5 kg/m2) and free of coronary heart disease, diabetes, and cancer at baseline (1957-8) and eligible (65+) for Medicare benefits from 1984 (the first year Medicare data were available for research use) to 2002. Height and weight were measured at baseline and annually thereafter until 1968. Participants were classified by average annual weight change: weight losers (<0 lb), moderate gainers (0-2 lbs), and heavy gainers (>2 lbs). Main outcome measures based on Medicare claims data were diabetes-related (International Classification of Diseases 9th revision code 250) inpatient and outpatient diagnoses and average annual charges (as surrogates for costs). RESULTS: The percentage of participants with diabetes diagnosis in older age was 13.7%, 15.0%, and 20.9% for weight losers, moderate gainers, and heavy gainers. After adjusting for race, education, and baseline age and body mass index, the odds ratio (95% confidence interval) for having a diabetes diagnosis was 1.15 (0.77, 1.71) for moderate gainers and 2.38 (1.44, 3.95) for heavy gainers compared to weight losers (p value for linear trend = 0.002). Adjusted average diabetes-related annual charges were $405, $567, and $1229 for the 3 groups respectively (p-trend = 0.03). CONCLUSIONS: In this cohort of primarily Caucasian men who survived to age 65 or older, weight gain in middle age was strongly associated with diabetes-related medical utilization and costs in older age. Our finding adds to existing evidence on the impact of obesity and weight gain on diabetes and has profound implications for Medicare expenditure given the rise in obesity and diabetes epidemic.