Population density is one of the key parameters for assessing the magnitude of population exposed to flood risk, and the better quality data we have, the better the assessment of risk. The aim of this study is to elaborate a high resolution spatially distributed population density grid, which estimates population at the commune scale with a reliability of over 90%. The novelty of the approach is population density estimation in a regular European grid, based on buildings vector data collected in the national topographic database. Using abductive reasoning in combination with statistics and spatial analysis, the authors extract approximate information about a population from the large scale topographic data. Moreover, linking the obtained population data with the cadastral data - by unique building identifier - allows for regular, quick and census survey independent updates of the population surface. A shortcoming of the approach is the issue of the possible existence of two houses per family, which leads to an overestimation of population. However, in the study area it affected only two of the total fourteen communes by 7-9%.