Abstract: To assess the quality of groundwater in Jilin City and analyze its causes, we select 8 indexes from the data of 34 water quality monitoring wells, including indexes of iron, manganese, ammonia nitrogen, etc., and use BP neural network to assess the water quality and compare the results with that of the grading method of attaching notes in Groundwater Quality Standard GBT14848-93. The assessment results of BP neural network are basically consistent with that of the standard, which proves that the results are credible and that BP neural network can be used to comprehensively assess groundwater quality. In BP neural network, the determination of weight parameters is not required, so it can avoid the subjective error. According to the zoning map of groundwater quality for plain area of Jilin city, IV and V class waters in the region account for a large portion whileⅡ and Ⅲ class waters are few. According to the spatial distribution maps of common factors, it is concluded that the main human causes for groundwater pollution are improper discharge of industrial waste water, smoky dust etc., leakage of sewage pipes, industrial waste and garbage and the excessive use of fertilizer and pesticide in farmland.