The change of groundwater table is one of the main characters of dynamic variation of groundwater. Two typical wells are chosen in Weibei dry highland of Shaanxi Province in China. Using the arailabe measured ground water table data from 2000to 2010, the paper carried out the simulation and prediction for groundwater table by BP neural network method. The results showed that the predicted values were in a good coincidence with the observed values. Therefore,this method can be used to predict the dynamic variation of groundwater.