In this study, groundwater vulnerability in the “Campus UANL” aquifer was assessed by applying a modified DRASTIC model. Pollution source mapping (P) and Management of water resources (M) were added to the DRASTIC parameters. A geographical information system was used for the spatial integration of different parameter maps. The effect of parameters on the resulting vulnerability maps was determined through a single parameter sensitivity analysis. The most influential parameters identified were topography, pollution sources mapping, net recharge, and soil type. Efficiency was validated by calculating the coefficient of determination (r²). Values for the DRASTIC-PM model were 0.63 for EC, 0.89 for Al³⁺, and 0.64 for Co²⁺, which improved to 0.68 for Co²⁺ after sensitivity adjustments. These validation results indicate that the modified model is more effective than the conventional DRASTIC model in identifying critical vulnerability zones within an urban metallurgical-industrial environment. The groundwater vulnerability assessed by applying the modified model revealed that the aquifer in the study has very high vulnerability. In contrast, the conventional model indicated that the aquifer exhibits high vulnerability. The results revealed a critical situation regarding groundwater quality, especially in the northeastern area. The DRASTIC-PM model proposed in this study provided a more accurate and detailed vulnerability zonation. This modified model can be implemented in zones where anthropogenic contamination is high, particularly in and around urban centers. However, the limitations of this study are related to the availability and quality of hydrogeological and geochemical data. The innovation of this study lies in the application of the DRASTIC-PM model, calibrated with metal ions and adapted to an urban-industrial setting, to assess the vulnerability of the “Campus UANL” aquifer. This approach allows for a more accurate evaluation of vulnerability in an area under high anthropogenic pressure, significantly improving the model's precision compared to conventional methods.