This paper divided the Liaohe River Basin into A, B, C, D and E 5 areas based on the 54 elements content of the surface soil by the neural network method, self-organizing feature mapping (SOM), and cluster analysis method, C means. The results showed that the soil in A District is fertile with high organic matter content, high and equal N, P and various trace elements content. The most elements content in the soil of B District are equilibrium and slightly lower than that of A District but higher than that of C District. The research has a significance for the human health, economic crop development and ecological environmental government of oft-com:office:smarttags" />