It is necessary to establish the proper indicators for soil quality assessment. In this paper a method with Peason two-tailed bivariate correlations and Cluster analysis was applied for the determination of soil quality index of urban forest soil. Urban soil samples at 0-20 cm depths were collected from the forests in Nanjing at different functional areas, and 24 soil indicators were analyzed in lab, finally 9 of soil indicators were selected to compose the minimum factor set for the fertility assessment of urban forest soil, including soil bulk density, clay content, dispersion coefficient, erodibility K values, water stable index, pH, organic matter, available potassium, and cellulase activity.