ABSTRACT Objectives Stepwise discrimination analysis method was adopted to establish function model of coronary heart disease（CHD） with SPSS software.And the clinical values of the relevant parameters by Quality Arterial Stiffness（QAS） were studied. Methods A total of 154 patients that doubtful coronary heart disease hospitailed in our cardiovascular department， undergoed coronary angiography and divided into non-CHD group（n=49） and CHD group（n=105）.All the patients were checked by carotid QAS，and the relevan parameters were recorded.Stepwise discrimination analysis method was adopted to establish function model of CHD .Results The difference of PWV、CC、DC、α、β、R、H was statistically significant between non-CHD group and CHD group.At last, Pulse Wave Velocity （PWV），Compliance Coefficient（CC），the reflected wave amplitude（H）and slope（R）came into the model，and the Fisher discrimination function was Y=-7.985-0.612CC+0.298PWV+0.028H+5.058R，The Bayes discrimination function was Y（Non-CHD）=-52.566+4.678CC+2.228PWV+0.629H+33.549R，Y（CHD）=-63.509+3.793CC+2.659PWV+0.669H+40.869R. The accuracy rate and false rate by retrospective inspection，Jackknife method inspection and External data validation were respectively 79% and 21%， 75% and 25%，81% and 19%.Conclusions The parameters including PWV、 CC 、R and H were closely related to coronary heart disease，and the established function model had an important reference value in discriminating the occurrence of CHD.