【Abstract】 Objective To establish a prediction formula for screening atherosclerotic renal artery stenosis (ARAS). Methods Clinical data of 892 patients who suffered from coronary artery disease (CAD) with renal artery angiography were reviewed. The clinical characters related to ARAS were analyzed and Logisitc regression was used to reveal the ratio of every clinical character. Based on the ratio, a prediction formula was established. Then sensitivity and specificity of this prediction formula was examined back to the patients. Results The patients who suffered from ARAS in CAD population accounted for 12.7%. The risk factors included age, BMI, Scr, hypertension history, diabetes mellitus (DM)，stroke and refractory hypertension. Age, BMI and Scr were used as continuous variables and other four as dichotomized variables. According to the prediction formula, the score was from 5.5 to 20.5. With the score increasing, the probability of ARAS increased as well. Conclusion This prediction formula is helpful to select the ARAS patients from CAD population, in order to provide them the evidence for further examination.