The paper considers the contribution of economic, residential development and residential location models to an understanding of the urban housing market and the residential location process. The cost-substitution approach is shown to possess little analytical power in relation to the location of both inner-city and suburban residents, because of the presence of a general positive price gradient throughout the city. Doubts are raised concerning the ability of residential location models to fully portray the urban market, owing to their descriptive nature and the map pattern problem. The problems facing the economic analysis of housing are also considered. A brief analysis of the applicability of Marxian concepts to the general area of housing is attempted. Areas of necessary research are also identified. GEOGRAPHY has never concerned itself deeply with the question of housing. Detailed discussion of the role of housing as an agent of segregation and dispersal, or, as Robson (1973) has defined it, as an institutionalfilter, remains sadly lacking. There still exists little work to bridge the aggregate portraits of the factorial ecologist, and the miniatures that depict the household decision-making process. It is no exaggeration to suggest that the workings of the housing market represent the key to the social structuring of the city. As Octavia Hill stated, 'you cannot deal with people and their homes separately' (quoted in Ambrose, I974). It is the intention of this brief paper to draw upon some of the available literature relating to the housing market, and to discuss the strengths and weaknesses of the work as it relates to the British context. This body of literature has been classified by Baxter and Anthony (I971) into three groups: (a) economic models of the housing market; (b) residential development models; (c) residential location models. In order that this taxonomy should create no confusion, the salient works within each group are listed below. Economic models: Alonso (I964), Harris (I966), Herbert and Stevens (1960), Kain (1962), Muth (1969), Reid (1962), Wingo (1961), Wolfe (1967). Residential development models: Chapin and Weiss (I968), Swerdloff and Stowers (1966). Residential location models: Baxter and Anthony (1971), Cripps and Cater (1971), Cripps and Foot (1969), Wilson (I970). It is not within the scope, nor is it the purpose of this paper to document all the above presentations, but rather to discuss the implications of the different approaches. In many cases the comments made will apply to all of them, for the assumptions are in general shared, and the methodology is similar.