In this paper, we construct a data set of Internet offer prices for flats in 48 large European cities from 24 countries. The data are collected in January - April 2012 from 33 websites, where the advertisements of flats for sale are placed. Using these data we investigate the determinants of the flat prices. Four factors are found to be relevant for the flats' price level: income per capita, population density, unemployment rate, and Gini index. The results are robust both to excluding variables and to applying two alternative estimation techniques: OLS and quantile regression. Based on our estimation results we are able to identify the cities, where the prices are overvalued, and those, where the prices are undervalued. This is a useful information that allows analyzing and comparing the housing markets in large European cities.