Abstract: In order to find analysis and assessment methods of phenotypic selection of officinal neem superior trees, and to empirically evaluate the officinal neem resources and to finally provide evidences and guidance in officinal neem breeding, 12 characters correlated with quality and yield in 32 primarily selected superior trees and control were tested. The results are as follows: among characters, at least one was significantly or extremely significantly correlated with another. Accumulative contribution rate of 4 common factors (with latent root >1) was 83.53%, their contribution rate was 31.33% (the first PC, seed shape), 22.98% (the second PC, azadirachtin content), 15.11% (the third PC, seed yield), 14.10% (the fourth PC, ratio of azadirachtin ingredient), respectively. And typical example of variables in analysis and assessment of phenotypic selection were: kernel weight (0.995 11), seed weight (0.956 76), seed length (0.780 12), content of azadirachtin A+B (0.866 67), content of azadirachtin B (0.819 34), content of azadirachtin A (0.680 84), ground diameter (0.939 04), seed yield per tree (0.722 82), ratio of azadirachtin B to A (-0.833 60). Mantel test showed that cluster analysis could reveal the phenotypic differences of 33 trees (r=0.81, p=1.00). Cluster groups could be categorized as: type of high azadirachtin content, type of full size of seeds and kernels, type of larger tree and high seed yield, and type of normal. Criterion of comprehensive assessment of primarily selected superior trees: "Superior" (with score ≥ 0.59), "Medium" (with score -0.190.59), "Poor" (with score ≤-0.19). According to this criterion, primarily selected superior trees were divided into 4 "Superior", 16"Medium", and 12"Poor" trees. With factor analysis and cluster analysis, analysis and assessment of phenotypic selection of officinal neem superior trees can be directly and easily performed and judged, the results could provide evidences and guidance in officinal neem breeding.