Aim: The aim of this study was to assess the predictive power of glycated hemoglobin (HbA1c) for future type 2 diabetes risk. Research Design and Methods: Six hundred eighty-seven subjectswhowere free of type 2 diabetes mellitus (T2DM) participated in the study. Each subject received a 75-g oral glucose tolerance test at baseline and 624 received a repeat oral glucose tolerance test after 3.5 ± 0.1 yr of follow-up.Anthropometric measurements, lipid profile, and HbA1c were measured during the baseline visit. Logistic multivariate models were created with T2DM status at follow-up as the dependent variable and other parameters as the independent variables. The receiver-operating characteristic (ROC) was used to assess the predictive discrimination of the various models. Results: HbA1c was a significant predictor of future T2DM risk (area under the ROC curve = 0.73, P < 0.0001). A HbA1c cut point of 5.65% had the maximal sum of sensitivity and specificity.Although the area under the ROC curve of HbA1c was smaller than the area under the ROC curve of both the 1-h plasma glucose concentration and a multivariate logistic model (including anthropometric parameters, lipid profile, and fasting plasma glucose), the addition of HbA1c to both the 1-h plasma glucose and the multivariate logistic model significantly increased their predictive power. Conclusion: Although HbA1c alone is a weaker predictor of future T2DM risk compared with the 1-h plasma glucose, it provides additive information about future T2DM risk when added to previously published prediction models.