This paper has conducted the national-scale assessment of flood disaster risk in China based on the theory of natural disaster risks and by means of the powerful spatial analytic tool of GIS, as well as the normalization and analytic hierarchy process. The precipitation of the current day and last three days, topographic elevation, topographic standard deviation, network of rivers and lakes, population density and GDP density respectively were extracted as assessment factors and risk assessment model was built for the national-scale flood disasters, which is similar to the mode of weather forecast model. It was applied for a real-time flood event from 18th Jun. to 16th Sep 2009 on daily basis, and validated based on the official report of disaster data. The rate of correct assessment (i.e. the risks and disasters occurs) is above 65% for counties as the assessment unit and above 84% for prefecture-level cities as the assessment unit. The validation result reveals that GIS-based national scale risk assessment model conforms basically to the actual disaster data.