Abstract: Objective Resistivity method is an important technique for detecting leakage in reservoir dams. However, when using surface resistivity methods to detect hidden dangers, issues such as blind zones near both abutments and insufficient resolution at depth arise, making it difficult to accurately identify weak areas across the entire dam section.Methods To address these limitations, this study proposes an integrated surface and borehole-based approach for dam seepage detection. Numerical simulations were conducted to analyze the distribution characteristics of surface-to-borehole resistivity under both lateral and vertical variations of seepage zones. Taking Shibi Reservoir Dam as a case study, a surface-to-borehole resistivity observation system was employed to acquire electrode current data, as well as near-surface, cross-borehole, and surface-borehole resistivity information. The reliability of the detection results was validated using borehole data, reservoir water level records, geological profiles, and borehole television imaging. Results The results indicate that as the preset seepage zone moves closer to the dam abutment, the deviation in the depth and central position of the anomaly identified by surface resistivity increases gradually. In contrast, surface-to-borehole resistivity tomography is less affected by the high-resistivity zones in the slope and shows higher consistency with the preset models. Surface resistivity is sensitive to shallow seepage anomalies but struggles to identify deep seepage zones. The surface-to-borehole resistivity inversion clearly reveals targets at different depths and offers strong capability in delineating geological boundaries. During the field test at Shibi Reservoir, surface resistivity indicated an extensive anomalous seepage zone within the dam body but failed to fully capture the seepage range near the abutment. Cross-borehole resistivity improved detection accuracy between boreholes within the dam body, yet its coverage remained limited. In comparison, surface-to-borehole resistivity inversion provided broader spatial coverage, effectively suppressed low-resistivity anomalies in the dam body, reduced the volumetric effect inherent in surface resistivity, and enabled precise delineation of low-resistivity anomalies in deep abutment contact zones.Conclusions This study provides a technical basis for targeted treatment of hidden hazards in reservoir dams and offers new insights into precise diagnosis and intelligent sensing of earth-rock dams.