The patterns and topographic complexity inherent in urban surfaces cause considerable variations in flow paths, and may influence the simulation outputs of urban stormwater management models (SWMMs). However, the influence of these variables in urban storm water management is largely unknown and so it is usually ignored, consequently reducing the accuracy of the simulation results. To examine how patterns of land use and overland flow routing options influence the simulation outputs of the SWMM, we chose an experimental site in Bazhong City. We developed three sub-area overland flow routing methods (outlet, impervious, permeable) for different precipitation intensities and land use patterns within the sub-catchment area module of the SWMM, and compared the simulation outputs. The results indicated that the simulated surface runoff for the outlet and impervious routing methods were the same, but were considerably different from those for the permeable routing method. Under the permeable routing method, the surface runoff decreased by 52% and the amount of rainfall that permeated almost doubled. The percentage of runoff routed by the permeable routing method had a significant impact on the simulation results. As the percentage of routed runoff increased, the directly connected impervious area (DCIA) gradually decreased and the unconnected impervious area (UIA) gradually increased, which led to a considerable reduction in the total runoff volume and a decrease in the surface runoff coefficient. Simultaneously, the amount of rainfall that permeated gradually increased, but the peak flow rate initially increased and then decreased. When the percentage of routed runoff reached 30% or 40%, the UIA was nearly equal to the permeable surface area, and the peak flow rate was at a minimum. The results from the present study showed that land use patterns and overland flow routing options influenced the SWMM simulation outputs. The results also suggest that for the effective management of urban stormwater, urban land-use patterns and overland flow paths should be optimized at a small scale, as this will reduce the risk of urban stormwater disasters. This approach will support the planning and development of sponge cities.