Abstract: In order to screen out the evaluation index of Cornus wilsonian seedling quality and classification criteria and the best way of dividing its quality grades. The morphological indicators like seedling height, ground diameter, fresh weight, dry weight were determined by taking one-year-old Cornus wilsonian actual seedlings as research objects, and then correlation analysis, principal component analysis, stepwise cluster analysis and average plus standard deviation for results were carried out. Correlation analysisi showed that fresh weight and dry weight were the important indexes evaluating the quality of seedlings of Cornus wilsonian, while seedling height and ground diameter were the main indexes. Combined with principal component analysis（PCA）, the first principal component candenote ground diameter was the biggest, and the second principal componment candenote height to diameter ratio was relatively big. Further, the ground diameter and seedling height were used as assessing the main indictors for the seedling quality of Cornus wilsonian. Classification of seedling quality by stepwise cluster analysis, in the case of three levels, seedling height of Grade Ⅰ was greater than or equal to 103 cm, ground diameter more than or equal to 0.59 cm; seedling height of Grade Ⅱ ranged from 93 cm to 103 cm, including 93 cm, ground diameter ranged from 0.50 to 0.59 cm, including 0.59 cm; seedling height of Grade Ⅲ was less than 93 cm, and ground diameter was less than 0.50 cm. The classification method of average ± standard deviation was used for classification, seedling height of Grade Ⅰ was greater than or equal to 100 cm, and ground diameter was more than 0.06 cm; seedling height of Grade Ⅱ ranged from 90 cm to 100 cm, including 90 cm, while ground diameter ranged from 0.50 cm to 0.60 cm, including 0.50 cm; seedling height of Grade Ⅲ was less than 90 cm, and ground diameter was less than 0.50 cm. There was a extremely little difference between the results obtained by the two classification methods, and the average ± standard deviation method for seedling quality classification of Cornus wilsonian was recommended to actual production.