This research work aim at investigating the behavior of the real estate market of the city of Porto Alegre, through the use of multiple variable statistical models, which explain the value formation of housing rents. The models developed were based on a sample of 504 apartments, from several different areas of the city, offered in the market during July 1992. The universe is all properties offered in that period. The models developed were based on a sample of 504 apartments, extracted of the total of 1819. In that apartments, a number of variables had their influence in the value of rent investigated, like building condition and proximity conditions, completed by information of public institutions. The data were tested by factor and regression analysis, with the compilation of econometric models (hedonic). in various aggregation levels. The analysis demonstrated what the market can to be explicated by this models, what present significative differences, according to data collection, but permit the obtainment of models for all dwellings and a best understanding of working of real estate market
housing market; rental housing; hedonic price models