The scientific evaluation of water quality change trends and pollution characteristics is of great significance to improving the current situation of water resources. The particle swarm optimization support vector machine based on principal component analysis (PCA-PSO-SVM) was used to conduct a comprehensive evaluation of groundwater quality in Xinzhou city, and the results were compared with those of a variety of traditional water quality evaluation methods. The evaluation results show that the water quality evaluation model based on PCA-PSO-SVM is more comprehensive and objective. The overall groundwater quality situation in Xinzhou city is generally good. The upper pore water quality is mainly class II, and the lower karst water is mainly class III. The water quality shows significant spatial differences. Compared with the traditional water quality evaluation method, the improved SVM algorithm compensates for the defects of the traditional method. The model structure is stable, and the accuracy and calculation efficiency are high, which makes the method worthy of promotion and application. The research results can provide a scientific basis and reference value for the water quality evaluation of groundwater-related projects.