Aim To investigate the relationship between brachial-ankle pulse wave velocity(baPWV) and aortic valve sclerosis(AVS) and the association of their severity, and the predictive value of AVS and baPWV in the diagnosis of coronary heart disease and three-vessel coronary artery disease among a group of hospital patients who underwent coronary angiography. Methods Performing coronary angiography, transthoracic echocardiography and the measurement of baPWV on 485 in-hospital patients, including 297 male cases and 188 female cases(average age 62.52±10.28 years). According to the results of echocardiography, patients were divided into AVS group and non-AVS group. They were also grouped by the number of valves involved and AVS scores. BaPWV was compared between each group. Patients were divided into 5 groups based on the results of coronary angiography:normal coronary, coronary atherosclerosis, 1-vessel disease, 2-vessel disease, 3-vessel±left main disease. To compare sensitivity, specificity, positive predictive value and negative predictive value of baPWV, AVS and their combination in the prediction of coronary heart disease and three-vessel coronary artery disease. To explore whether combining baPWV and AVS can improve the predictive value. Results There were 305 cases in AVS group and 180 cases in non-AVS group. BaPWV of AVS group(1536.57±285.06 cm/s) was significantly higher than non-AVS group(1484.92±241.75 cm/s, P<0.05). With different number of valves involved, baPWV value in 0-valve group was 1484.92±241.75 cm/s, in 1-valve group was 1499.61±259.57 cm/s, in ≥2-valve group was 1593.55±313.07 cm/s, and it had statistically significant difference(P<0.05). With the AVS scores, baPWV value in 0 score group was 1484.92±241.75 cm/s, in 1 score group was 1500.23±271.70 cm/s, in ≥2 scores group was 1586.16±296.26 cm/s, and it had statistically significant difference(P<0.05). AVS was superior to baPWV both in the prediction of coronary heart disease and three-vessel coronary artery disease. AVS had a positive predictive value of 91.1% in predicting coronary heart disease. BaPWV and AVS both positive had a high specificity in the prediction of coronary heart disease and three-vessel coronary artery disease(85.5% and 66.9%), better than separate prediction. High sensitivity appeared in the predicition of coronary heart disease and three-vessel coronary artery disease(90.4% and 95.3%) when either of baPWV and AVS was positive. Although AVS score ≥2 had a low sensitivity(33.6%) in the prediction of coronary heart disease, yet its specificity and positive predictive value were rather high(97.3% and 97.7% respectively). Conclusion BaPWV in AVS group was higher than in non-AVS group. With different number of valves involved, baPWV value in ≥2-valve group was significantly higher than baPWV value in 0-valve group and 1-valve group. With the AVS scores, baPWV value in 2 score group was significantly higher than baPWV value in 0 score group and 1 score group. With the combined application of baPWV and AVS in the prediction of coronary artery disease, both positive had higher specificity, either positive showed higher sensitivity. In the prediction of coronary heart disease, AVS score ≥2 showed low sensitivity but it had very high accuracy.