In this study， establish an efficient method to assess the eating quality of cooking-soybean by principal component analysis（PCA） combined with partial least squares（PLS）. Taking hardness， gumminess and sensory core of post-cooked-soybean as dependent variables， main factors influence on eating quality of cooking-soybean were screened for principal component analysis by the correlation coefficient and the variable importance value of the projected （VIP） of each component conducted by PLS model with the indexes of physicochemical properties， water absorption rate and swelling rate after soaking of soybean as variations. Results showed that： calcium content， 7S globular， soluble sugar content and swelling rate were screened out by PLS， which had a major impact on the texture quality of post-cooked soybean. Four principal components obtained by PCA， which could reflect most details on eating quality of cooking-soybean with a cumulative contribution rate of 73.374%. The hardness， gumminess， 7S， 11S/7S and total Ca were the main factors affecting the eating quality of cooking-soybean on the basis of contribution rate of each principal component. The quality of different varieties on eating quality of cooking-soybean was comprehensively evaluated by using the principal component comprehensive score model. Five suitable soybean varieties suitable for cooking and processing were selected as follows： Heinong85， Hefeng55， JD149， Suinong42 and Dongnong251.