The study deals with urban economic residential location modelling of Budapest, Hungary, using a pragmatical neural network approach. First the analysis pertains to the city level spatial housing market structure through identifying a number of factors related to housing supply, demand and price. After that, the analysis is zoomed in to involve the dynamics of two selected inner city neighbourhoods: the middle-parts of the districts VIII (JÛzsefv·ros) and IX (Ferencv·ros) respectively. These areas in the south-eastern part of the inner city have both received attention as subjects for substantial rehabilitation in recent decades. The two districts are adjacent, but different: the former is stigmatized in all discourse although it comprises a great variety of microlocations and also housing stock; the latter district in turn is more homogeneous, partly gentrified area and undoubtedly the most dynamic neighbourhood in the city with the best quality apartments on the Pest side. Earlier results (reported elsewhere) based on a small set of individual mortgage valuations suggested the Budapest housing market to be spatially and functionally extremely fragmented, as a mosaic of various house types, age-categories and price-levels, as well as micro-locations, could be identified. A follow-up analysis of street- and district-wise aggregated data with the same neural network modelling techniques, namely the self-organizing map (the SOM) and the learning vector quantification (the LVQ), shows further evidence about the balance between physical and socio-demographic characteristics and house price levels. The new results are synchronised with the earlier findings: also here the administrative district is found less important determinant of the market position of the dwelling than its price, type and immediate vicinity. After that the case-study areas were modelled using the SOM for each crosssection during the observed six-year period from 1997 to 2002. Two types of dynamics could be identified, as the housing market dynamics in the selected ëslice' of the urban area has differentiated the micro-locations in terms of house price escalations on one hand, and in terms of changes in the urban structure on the other. Above all the evidence reveals how the housing market development is related to the most localized processes of social and physical upgrading taking place in an urban setting.