OBJECTIVETo study the association between vitamin D status and the risk of incident impaired fasting glucose (IFG) and diabetes in a population-based cohort of diabetes-free subjects.RESEARCH DESIGN AND METHODSIn a historical prospective cohort study of subjects from the Clalit Health Services database, which includes information on nearly 4 million people, diabetes-free subjects aged 40–70 years with serum 25-hydroxycholecalciferol (25-OHD) measurements available were followed for 2 years to assess the development of IFG and diabetes in five 25-OHD subgroups: ≥25, 25.1–37.5, 37.6–50, 50.1–75, and >75 nmol/L.RESULTSThe baseline cohort included 117,960 adults: 83,526 normoglycemic subjects and 34,434 subjects with IFG. During follow-up, 8,629 subjects (10.3% of the normoglycemic group) developed IFG, and 2,162 subjects (1.8% of the total cohort) progressed to diabetes. A multivariable model adjusted for age, sex, population group, immigrant status, BMI, season of vitamin D measurement, LDL and HDL cholesterol, triglycerides, estimated glomerular filtration rate, history of hypertension or cardiovascular disease, Charlson comorbidity index, smoking, and socioeconomic status revealed an inverse association between 25-OHD and the risk of progression to IFG and diabetes. The odds of transitioning from normoglycemia to IFG, from normoglycemia to diabetes, and from IFG to diabetes in subjects with a 25-OHD level ≤25 nmol/L were greater than those of subjects with a 25-OHD level >75 nmol/L [odds ratio 1.13 (95% CI 1.03–1.24), 1.77 (1.11–2.83), and 1.43 (1.16–1.76), respectively].CONCLUSIONSVitamin D deficiency appears to be an independent risk factor for the development of IFG and diabetes.