Study region: The study area focuses on 15 small watersheds in Shanxi Province, China, spanning the Yongding, Ziya, and Zhangwei river systems. These watersheds are located in a transitional zone from semi-arid to semi-humid climates, with complex topography and extremely sparse hydrological data and intensive human activities. Study focus: This study evaluates flood hazard conditions across the 15 watersheds by integrating hydrological methods with the Integrated Flood Modeling System (IFMS). Flood hydrographs and inundation dynamics were simulated for five return periods (5-, 10-, 20-, 50-, and 100-year). Comparative analyses were performed across three hydrological zones and three land use types to quantify the coupled effects of natural and anthropogenic drivers on flood hazard characteristics. New hydrological insights for the region: The Shizi, Mian, and Yangwu River basins exhibit the most severe flood hazards, with maximum inundation depths exceeding 2.0 m, inundated areas larger than 3.0 km2, and affected populations over 1200 persons. Basins in Hydrological Zone II (semiarid) show higher mean inundation depths (1.69 m) than those in the more humid Zone III (1.27 m), which is mainly attributed to sparse vegetation and a high impervious surface fraction (exceeding 70%). Watersheds dominated by Cropland-Built-up land (CBUL) show the highest vulnerability, while Forest-Grassland (FGL) and Water-Rich (WR) types exhibit significantly lower flood hazards. Reservoirs in Water-Rich basins notably attenuate flood impacts. These results reveal that human-altered land use can override climatic gradients to dominate flood hazard dynamics, providing a scientific basis for spatially targeted flood mitigation strategies in data-scarce regions.