Aim To evaluate the value of peripheral atherosclerosis,including ocular fundus atherosclerosis(OFA),brachial-ankle pulse wave velocity(BaPWV)and ankle-brachial index(ABI),in predicting the extent of coronary artery lesions according to coronary angiography(CAG). Methods 472 consecutive patients who had taken CAG were studied.All patients were underwent OFA,BaPWV,ABI,color doppler echocardiography(UCG) examination and serum factors evaluation in addition to history collection,which were divided into two groups according to the result of the CAG.Group one was non-coronary artery disease(CAD),which has a normal result of CAG;group two was CAD group,which has an abnormal result of CAG(coronary lesions≥50%).CAD group was divided into two sub-groups according to the lesions of the vessels.Sub-group one has single or two vessels lesion;sub-group two has three or more vessels lesion or left main lesion.A Logistic regression model for predicting coronary lesions on the basis of OFA and other risk factors was obtained.A receiver operating characteristic(ROC) curve was used to assess the performance of the Logistic regression model. Results On the one hand,the CAD group as the dependent variable and the risk factors as the independent variable: diabetes mellitus,BaPWV>14 m/s,typical symptom of chest pain,two or more grade of OFA and UCG showing abnormal motion of ventricular were applied into the model.Only typical symptom of chest pain had a statistical significance to the CAD(OR=7.220,ROC=0.736±0.024,P=0.000).But other factors such as two or more grade of OFA(OR=3.969,P=0.056),BaPWV>14 m/s(OR=2.846,P=0.061),diabetes mellitus(OR=2.769,P=0.063) and UCG showing abnormal motion of ventricular(OR=2.513,P=0.070) had no statistical significance to the CAD.On the other hand,the sub-group as the dependent variable and the risk factors as the independent variable: diabetes mellitus(OR=3.210,P=0.054),two or more grade of OFA(OR=3.014,P=0.062),BaPWV>14 m/s(OR=2.860,P=0.063) were applied into the model,all these factors had no statistical significance to predict the three or more vessels lesion.However,if any two of them were combined,two or more grade of OFA and BaPWV>14 m/s had a statistical significance to predict the three or more vessels lesion(OR=6.428,ROC=0.736±0.024,P=0.041);If all the three factors were combined,they had a very statistical significance to predict the three or more vessels lesion of the coronary(OR=43.466,ROC=0.736±0.024,P=0.000). Conclusion The typical symptom of chest pain is positively correlated with the result of CAG,which is useful for predicting the coronary artery disease.Diabetes mellitus,two or more grade of OFA and BaPWV>14 m/s are positively correlated with the vessels lesion of the coronary and these three factors combination is powerful for predicting the three or more vessels lesion of the coronary.