The Topographic Index Model (TOPMODEL) has been implemented into land surface models (LSMs) to improve modeling of hydrological process components. In recent years, great effort has been made by many researchers for adjusting some of the assumptions contained in the classic TOPMODEL, which will broaden its application scope. For example, the general power law has been used to describe the variation of saturated hydraulic conductivity with depth as an extension of the classic TOPMODEL. Known as power law TOPMODEL, its derivation is based on the assumption of spatially uniform land surface. In this paper, the power law TOPMODEL is extended to a spatially inhomogeneous land surface with spatially non-uniform saturated hydraulic conductivity at the ground surface (K₀), effective soil depth (m), and water recharge rate to the ground water (R). In addition, numerical experiments with assumed spatial variable patterns of K₀, m, and R are conducted to evaluate the hydrological effects of the spatial heterogeneity. The main conclusions are as follows: 1) The assumed spatial distributions of K₀, R, and m affect daily surface runoff, daily baseflow, and daily total runoff. Among these, the spatial distribution of m significantly increases daily surface runoff and flood discharge. 2) The assumed spatial distributions of K₀, R, and m do not significantly affect the total runoff and evaporation averaged over many years, although they change the partition of total runoff between surface runoff and baseflow. In particular, the spatial distribution of R₀ enhances averaged surface runoff of many years and significantly reduces the averaged baseflow during the same period. The other two variables change surface runoff and baseflow less than does the assumed spatial distributions of R₀.