Rapid urbanization has greatly changed the underlying surface, resulting in frequent floods and waterlogging in urban areas and causing enormous economic damage and loss of life. Of the innovative storm water management strategies, green infrastructure (GI), which uses vegetation, soils, and natural processes to manage storm water and create healthier urban environments, has proved to be an effective measure to mitigate urban flooding. The runoff reduction effectiveness of green infrastructure facilities under different scenarios can be simulated by mathematical models. However, complex models like the storm water management model are difficult for planners and managers to operate and apply. Furthermore, they are inadequate in demonstrating the runoff reduction mechanisms of green infrastructure. Although the performance of green infrastructure in mitigating urban flooding has been extensively investigated, few studies have attempted to examine and compare runoff reduction effectiveness between integrated green infrastructure and single green infrastructure facilities under different storm recurrence periods. In this study, a community-scale simulation model based on water balance and urban hydrological processes was developed to quantify the reduction effect of green infrastructure on storm water runoff. A typical community in Beijing was selected as a case study to assess the reduction effectiveness of different green infrastructure configurations, based on the volume and peak flow of storm water runoff under 1-year and 5-year storm events. Four scenarios of green infrastructure configuration were considered, namely: converting to concave green land; constructing a storage pond; converting to porous brick pavement; and combining the previous three measures. Field-monitored runoff data of two rain events were used to validate the model. The validation results yielded determination coefficients of 0.68 and 0.71 respectively for the two rain events, while the Nash-Sutcliffe efficiencies were 0.99 and 0.96 respectively, indicating that model performance was satisfactory and reliable. For the scenario of concave green space with 5 cm depth, runoff volume was reduced by 8.23% and 23.30% and peak flow was reduced by 20.31% and 29.11% respectively for 1-year and 5-year storm events. For the 300 m³ storage pond scenario, runoff volume was reduced by 84.90% and 20.97% respectively for 1-year and 5-year events, while peak flow was reduced by 88.99% and 0.10% respectively. For the scenario in which 50% of impervious surface area was modified to porous brick pavement, runoff volume was reduced by 46.51% and 38.52%, and peak flow was reduced by 39.96% and 35.48% respectively for 1-year and 5-year events. These results indicate that each of the three facilities showed good runoff reduction effectiveness. With increased rainfall intensity, the reduction effectiveness of concave green space, storage pond, and porous brick pavement was slightly enhanced, decreased, and relatively stable respectively. The integrated green infrastructure configuration scenario showed significant reduction effects, with 100% reduction of runoff generated by the 1-year storm; runoff volume and peak flow were reduced by 75.47% and 64.52% respectively under conditions of the 5-year storm, as well as showing increased rainwater infiltration and harvesting for utilization. Therefore, the integrated green infrastructure configuration is among the optimal strategies for storm water runoff reduction and rainwater resource utilization in communities.