In the context of climate change and urbanization, increasing pluvial flood disasters leads to severe losses and impacts of urban inhabitants, properties and functions, particularly in the interruption of public services (e.g. urban transportation and emergency response). Therefore, the emergency management and risk adaptation of urban pluvial flooding have been one of the hotspots in natural hazards research. The city center (East Nanjing Road- Central Business District) of Shanghai, China has been selected as a study area because it exhibits enhanced consequences of pluvial flooding. A simple 2D hydrodynamic model (FloodMap-HydroInundation2D), which couples hydrological processes with surfacewater flow modeling in urban environment, was used to derive pluvial flood inundation associated with urban road network. Furthermore, GIS- based spatial analysis (network analysis and cost distance algorithm) was employed to evaluate the transportation conditions and emergency responses (accessibility and service coverage) of critical public service sectors (Medical Treatment and Public Security) under multiple scenarios of rainfall magnitudes (1 in 10 year and 1 in 100 year return periods) and flood thresholds (30 cm and 50 cm). The results show that: (1) at normal (no rainfall or less than 1 in 1 year rainfall) condition, urban emergency response could cover almost the total (more than 99%) area in 10 minutes or less, among which ambulances and police vehicles could reach the majority (82% and 87%) of the study area within 5 minutes. (2) During 1 in 10 year rainfall scenario, most of the urban roadways was predicted to be accessible and more than 90% of the city center would be reachable in 10 minutes. The response areas of large or medium emergency vehicles with better performance of water-wading (above 50 cm) would even be close to the service coverage at normal condition. (3) Under 1 in 100 year rainfall scenario, as the rainfall intensity significantly overwhelm the design capacity of storm sewer drainage system, majority of the urban road network would be highly inundated. Over half of the study area could be completely inaccessible within 10 minutes, according to 30 cm threshold. By contrast, the emergency service area would largely increase to about 93% of the city center when flood restriction subject to 50 cm threshold. This study suggests that the framework developed for coupling high resolution pluvial flood modeling with urban emergency response assessment, is proved to be practical and effective and will provide a scientific basis for urban flood mitigation and smart emergency management.