Abstract: In order to shorten the breeding cycle of Chinese fir (Cunninghamia lanceolata) clone, and select good clones with stronger tolerance to barrenness, in the light of 31 clones, artificial design nutrient environments, the experiment was carried out according to two factors randomized blocks, 8 individuals (in 8 cups) in each plot. After the experiment finished, the height and diameter at base of seedling were measured on entire experiment. The other traits such as the aboveground biomass, the underground biomass, the total biomass, the root-shoot ratio and the entire P content of individual have been measured according to sampling method. The results of statistical analysis showed that: (1)There were obvious differences among the clones in all the traits besides root-shoot ratio; (2)Obvious differences in all the traits existed under different nutrient stress and different sites conditions among different clones; (3)The interaction of clone × nutrient coercion processing was complex; (4)The research traits showed a certain variation and repeatability over intermediate level. Based on the variance analysis (fixed model), the PCA selection index under different situations was derived and obtained. Using certain selected intensity, 13 outstanding clones were evaluated from the entire experiment. Among them 7 clones were anti-phosphorus and the anti-potassium lack (happy nitrogenous fertilizer) clones, 7 clones were anti-phosphorus lack (happy nitrogen potassium fertilizer) clones, 8 fine clones were selected under normal nutrient management (happy nitrogenous, phosphorus, potassium), there were 5 fine clones with good performance, and simultaneously very stable under all kind of conditions. These clones may be applied to field tests.