Groundwater contamination prediction could be achieved through the numerical simulation model of groundwater contaminant transport.In order to analyze the influence of the uncertainty of parameters in the model, Monte Carlo method was used to analyze the uncertainty of the model output in this paper,In order to reduce the complexity of the numerical simulation model, the sensitivity analysis method was used to select the random variables in the model.To reduce the calculated load generated by the numerical simulation model called repeated, set up Krig surrogate model instead of the numerical simulation model to complete the simulation process. The results showed that the probability density function integral could be used to estimate the groundwater contamination risk and the contamination concentration range under different confidence level. The distribution of contaminant plume could estimate different level pollution cover acreage of the study area and the classification contamination risk early warning could estimate the contamination plume distribution of the study area in different contamination risk; Groundwater contamination risk warning, which was based on the uncertainty analysis of groundwater contaminant transport numerical simulation could forecast the groundwater problem more objectivily.