Abstract: Objective To evaluate the effect of different cut-point on inpaired fasting glucose(IFG) detection in community elderpeople.Methods A total of 1645 elderly people were followed-up for detection of IFG,cardiovascular disease,and diabetes mellitus from 2004 through 2008.Results The prevalence rate of IFG was 22% and 11% according to the criteria of American Diabetes.A ssociation(ADA) in 2003 and 1997,respectively.The prevalence of hypertension and coronary heart disease diagnosed with the 1997,ADA criteria was higher than that of with 2003 ADA criteria.In 2008,the incidence rate of IFG was 7.3% and 3.6% according to the ADA criteria of 2003 and 1997,respectively,and there was a significant difference in IFG outcomes behv een subjects diagnosed with different criteria There was no significant difference in incidence rates of hypertention and coronary heart disease behv een subjects with and without IFG based on differentADA criteria.Conclusion The IFG is a predictor for diabetes mellitus and different criteria of IFG could result in different IFG prevalence in a sam a population.