The characteristic feature of real estate market is local character, which marks that the conditionings of local market have very large influence on housing prices and dynamics of their changes. The study of changes housing prices as a result of outflow of time should, therefore to take into account both general trend and spatial differentiation, resulting with values of localization. The paper presents possibility of use geographically weighted regression to analysis of dynamics of changes housing prices on the area of Olsztyn in years 2003–2009. The simple global model in this aim was built for whole area of analysis and then the models of geographically weighted regression taking into account the spatial dependences weights. The conducted analysis confirmed the usefulness of the GWR method, as tool of spatial analysis of real estate market.

5-14