Establishing a simple method to predict the possibility of suffering diabetes by using some data that easy to acquire at home.By this way，we can detect the undiagnosed patients as earlier as possible.Methods:A total of 109 first-relationship relatives of type 2 diabetes participated in the test. We measured their blood sugar，age，neck circumference(NC)，waist circumference，hip circumference（WC），height，weight, and calculated their waist-hip ratio(WHR)and body mass index (BMI).Correlation，Chi-squared test and Logistic regression were used in the analysis of the relations between these data.Results:The prevalence rate of diabetes is 12.5% and 33.3% respectively before and after 55 years of age.The prevalence rate of diabetes is 27.9% and 3.6% respectively in the participants of obesity and non-obesity.The percentage of individual with normal FBG and abnormal PBG is 23.9%.Both correlation coefficient of NC / WC and NC/BMI are ＞0.7.Conclusion:The prevalence rate of diabetes increased significantly after 55 years of age.The prevalence rate of diabetes increased significantly in the obese people.Only measuring FBG would miss more IGT population and diabetic patients.NC has close correlations with BMI, WC，and is an appropriate indicator to reflect the degree of obesity.