Abstract: Thin coal seam mining is an important issue for the coal mining in China, thin coal seam mining method selection become more and more important. The geological conditions, management and equipment decide to the optimization selection of thin coal seam mining method base on the characteristics of thin coal seam mining, an optimization selection expert system(ES) of thin coal seam mining method based on neural networks (NN) is built. This system uses the self-adaptation algorithm to train the network. Finally, the mining method, yield and work efficiency can be predicted by this system.