Abstract: Analyzing the variation characteristics of groundwater flow field is crucial for monitoring and early warning of water inrush disasters. This study employed self-potential (SP) method in laboratory experiments to investigate water inrush processes during tunnel excavation through fault zones. The temporal evolution of SP signals was recorded, and the correlation between SP values and water inflow rates was statistically analyzed to elucidate the SP response mechanism during water inrush events, with numerical simulations conducted for validation. Key findings include: (1) SP variations become more pronounced closer to the water inrush section; (2) The rate of SP change exhibits high sensitivity to water inrush processes, effectively reflecting seepage velocity variations; (3) A significant negative correlation exists between SP values and water inflow rates (increased inflow corresponds to decreased SP); (4) SP signals demonstrate excellent response to both the location and magnitude of tunnel water inrush.