Abstract: Based on the Artificial Neural Network technology, this paper analyzed the height prediction method of the water conducted zone for mining under Xiaolangdi reservoir, selected lithology、compressive strength、types、thickness and the scale of mudstone in the overburden rock, obliquity and mining thickness of the coal seam as the main influence factors to establish the water conducted zone height prediction model, which forecasted the results effectively. The achievements of this study provided Xin'an coal mine with some important parameters and technological supports for rational design of the mining ways and means.