Abstract: The large number and concentrated distribution of pollution sources in coal-related industry agglomeration area can easily induce groundwater pollution, and it is of great significance to carry out the risk assessment of groundwater pollution to protect the groundwater environment in such areas. Taking a coal-related industrial agglomeration area in Shanxi Province as study area, the DRASTIC and PLEIK models were used to evaluate the vulnerability of pore water aquifers and karst water aquifers in the study area, respectively. Analytic Hierarchy Process (AHP) was used to determine the weights of the indicators in the PLEIK model, and the integrated loading of pollutants and the vulnerability of groundwater were used to characterize the risk of groundwater pollution in the study area. At the same time, combined with the water quality level of groundwater sampling points in the study area, the Random Forest (RF) classification algorithm was utilized to construct the groundwater pollution risk classification prediction method, and the evaluation results were compared with those of the superposition index method. The results show that: ① the loading of pollution sources in the study area is high, and the proportion of high pollution source loading area is about 26.73%, which is related to the characteristics of the more concentrated distribution of pollution sources in coal-related industry agglomeration area, and when quantifying the loading of pollution sources, the superposition effect of multiple pollution sources is obvious; ② the comprehensive vulnerability of the groundwater in the study area is dominated by the medium grade, and the proportion of medium vulnerability area is about 82.59%, with the pore-water aquifer high vulnerability area The high vulnerability zone of pore water aquifer is mainly located in the eastern and southeastern part of the study area, and the high vulnerability zone of karst water aquifer is mainly distributed in the karst exposed area north of Fenhe River; ③ the areas of low, medium and high risk zones of groundwater in the study area based on the superposition index method are 3.55%, 59.67% and 36.77%, respectively, and the consistency rate with the water quality level of the actual sampling points is 75%, and the risk of groundwater pollution predicted by the use of RF The predicted risk of groundwater contamination was dominated by low risk, and the correct rate between the grading results and the water quality of the actual sampling points was 97.7%, which improved the accuracy of the results calculated by the stacked index method by about 22.7%. The evaluation results are intended to provide a basis and reference for groundwater pollution control in the study area.