Abstract: Identifying the sources of mine water gushing is fundamental for preventing water hazards in coal mining areas, enhancing the economic efficiency and effectiveness of prevention measures. Taking a mining area in the southern margin of Ordos Basin as the research area, 11 mine water samples were collected before and after the mine water gushing events. Hydrogeological comprehensive analysis was employed to preliminarily identify the water sources, and a mathematical model was developed using groundwater chemical compositions. The sources of water gushing were quantitatively analyzed using statistical methods such as cluster analysis, factor analysis, multiple regression analysis, and canonical correlation analysis, the analysis results are mutually verified with the actual exploration results. The main sources of water gushing in the study area include Zhiluo Formation water, Luohe Formation water, and Quaternary water. Among these, Luohe Formation water and Quaternary water were primarily recharged by surface water, which infiltrated the mine through fractures. Cluster analysis showed that the first inrush water sample was grouped with Zhiluo Formation water, Quaternary water, and surface water, while goaf water formed a separate cluster. The chemical compositions of the second and third gushing water samples were similar to those of goaf water, the mass concentration of sulfate ions (SO₄²⁻) and magnesium ions (Mg²⁺) is higher. Factor analysis revealed that Zhiluo Formation water and goaf water contributed significantly to the water gushing, with contribution rates of 0.989 and 0.988, respectively. Multiple regression analysis indicated that surface water and Quaternary water had the most significant impacts on water gushing. Canonical correlation analysis further confirmed the strong relationship between gushing water samples and surface water or Quaternary water with regression coefficients of 2.764 and −3.169, respectively. The main sources of water gushing in the study area were identified as surface water, Quaternary water, and Zhiluo Formation water, with the findings highly consistent with actual exploration results. Therefore, integrating multivariate statistical analysis techniques can achieve quantitative identification of water gushing sources.