On the housing market literature there is a consensus to adopt the existence of submarkets and the recognition that the urban housing market should be portrayed as a set of distinct and inter-related submarkets. It is important to understand how the metropolitan areas are divided into housing submarkets for several reasons, including: increasing the accuracy of statistical models that are used to estimate the price of houses, improving the ability of lenders and investors in define the financial risk; and provide information to the buyer to find a residence. Segmentation can be attributed to individual characteristics of households, attributes of the site and the characteristics of the neighborhood. In this context, spatial econometrics is an important tool as it considers the spatial location of housing units. Therefore, this study aims to determine the housing market segmentation in S„o Paulo using property data from the last ten years (3,590 vertical condominiums). The program GeoDa will be used, since it allows doing statistical tests from geo-referenced data to detect the influence of spatial effects in hedonic models of house prices.