Abstract:Principal component analysis（PCA）and stepped-linear regression（LR）were used to study the properties of the fruit，including fruit weight，pulp hardness，fruit shape index，total soluble solids（TSS）content，and titratable acid content（TA），stone cell content，vitamin C（VC）content，the ratio of sugar to acid 29 indexes were analyzed and evaluated.The results showed that 7 principal components were extracted from 9 pear cultivars by principal component analysis，and the cumulative variance contribution rate was 97.86%，which could summarize the total information of these traits to a certain extent.The comprehensive quality prediction model of pear was established by factor analysis and stepwise regression.According to the scores of different principal components，seven simplified indexes were selected for comprehensive evaluation of different pear varieties，included weight per fruit，aroma，hardness，diameter of core/diameter of fruit，ratio of solid to acid，good fruit percentage 1，L value and b*value，which could provide theoretical reference for scientific evaluation of comprehensive quality of different pear varieties，production and promotion，as well as fruit storage and transportation.