Abstract: Evaluating effectively the loss caused by flood disaster has theoretical significance and practical value to flood-fighting, emergency rescues, flood relief and post-disaster compensation. However, the regional socioeconomic data affected by flood is usually collected in administrative region in the current flood loss evaluation method. The flood affected area and administrative area boundary do not necessarily match. The accuracy and practicability of flood loss evaluation based on loss rate are not soundly reliable. As an international IT research hotspot, Spatial Information Grid provides a new approach for flood damage assessment. This paper studies the flood loss assessment based on spatial information grid of socio-economic data. First, we establish a flood zone socio-economic data spatial information grid, including flood inundation zone special character mesh, grid cells use ranks number as the two key fields superimposed grid superimposed and flood damage assessment of spatial information grid. Followed by analysis of flood special character of the network division and attribute selection, considering the computational speed and convenience of other factors, taken rule quadrilateral partition method to divide the flood special character mesh, chose elevation, water depth, water velocity, water pollution degree of flood special Network personality attributes and based on GIS technology to generate a flood special character mesh. Then through the analysis of different disaster bodies submerged under different circumstances the loss difference, the loss rate is established based on flood damage assessment model. Finally, ArcEngine developed using C# and socio-economic data based on the realization of the spatial information grid flood damage assessment systems and use of heavy rainfall Poyang Lake county flood damage assessment data were calculated. The example shows that, based on socio-economic data for spatial information grid flood damage assessment methodology applied not only spatial information grid technology and flooded the area with flood damage to the region’s socio-economic data organically combined, so that it can quickly and accurately ground for inundated areas to assess the economic loss calculation. But how to select the appropriate grid scales and combining hydrological dynamics model enables more accurate flood special character network is the next step needs to continue to improve the research content.