Smoking is associated with low serum carotenoid concentrations. Prospective studies have found lower diabetes risk among persons with high-carotenoid diets. Whether diabetes risk is low in the rare smoker who has high serum carotenoid levels is unknown. The authors investigated the interaction of serum carotenoid concentrations and smoking with diabetes mellitus in 4,493 Black and White men and women aged 18-30 years in the Coronary Artery Risk Development in Young Adults (CARDIA) Study. The authors assessed 15-year (1985-2001) incident diabetes (148 cases), insulin concentration, and insulin resistance (homeostasis model assessment) in smokers and nonsmokers according to baseline levels of serum alpha-carotene, beta-carotene, zeaxanthin, beta-cryptoxanthin, and lycopene. Diabetes incidence was inversely associated with the sum of carotenoid concentrations in nonsmokers (per standard deviation (SD) increase, relative hazard = 0.74, 95% confidence interval: 0.55, 0.99) but not in current smokers (relative hazard = 1.13, 95% confidence interval: 0.83, 1.53) (p for interaction = 0.02). Similarly, year 15 insulin and insulin resistance values, adjusted for baseline levels, were inversely related to sum of carotenoids only in nonsmokers (per SD increase in insulin level, slope = -0.46 (p = 0.03); per SD increase in insulin resistance, slope = -0.14 (p = 0.01)). In CARDIA, higher serum carotenoid concentrations are associated with lower risk of diabetes and insulin resistance in nonsmokers but not in smokers.