Abstract: Seleting thirteen factors from four categories including geological structure, hydrogeological characteristics, floor aquifuge features and mining activities as discrimination index for floor water inrush, taking water inrush cases of typical coal mines happened in recent years as learning sample, distance discriminant analysis model for coal mine floor water inrush is established according to Fisher discrimination method. The results show that the the predicted results of model are consistent with the actual situation, and the misjudgement rate got from back substitution estimation method is small. So selelting four categories thirteen factors to reflect the situation of coal mine floor water inrush and using Fisher liner discrimination method to predict coal mine floor water inrush have good practicality and effectiveness.