To explore the association of dietary vitamin intake from food and/or supplement with mortality in US adults with diabetes.This prospective cohort study was conducted on 5418 US adults with diabetes from the National Health and Nutrition Examination Survey 1999-2018. Vitamin intake from food and supplements was estimated via dietary recall. Sufficient intake from food or food + supplement was defined as ≥ estimated average requirement (EAR) and ≤ tolerable upper intake level (UL), insufficient intake, < EAR; and excess intake, > UL. Medium supplementary intake was classified as > median level and ≤75th percentile; low intake, ≤ median level; and high intake, >75th percentile, as reported by supplement users.A total of 1601 deaths occurred among the participants over a median follow-up of 11.0 years. Cox regression analysis of the single-vitamin model demonstrated that sufficient vitamin A and folate intake from food and food + supplement and medium vitamin A and folate intake from supplement; sufficient riboflavin, niacin, and vitamin B6 intake from food and food + supplement; and sufficient thiamin and vitamin E intake from food + supplement were significantly associated with reduced all-cause mortality (all p < 0.05). In the multivitamin model, sufficient vitamin A and folate intake from food and food + supplement, medium vitamin A and folate intake from the supplement, and sufficient niacin intake from food and food + supplement were inversely associated with mortality (all p < 0.05).Vitamin A and folate intake from food or supplement and niacin intake from food were significantly associated with reduced mortality in US adults with diabetes.© 2023 John Wiley & Sons Ltd.