This article proposes an alternative method to Markovian approaches through the housing systems analysis (ASHA) model. Its objective is to simulate the impact of the housing stock on population redistribution at a given scale and duration, by modelling the processes whereby residential mobilities are linked, initiated either by a change in the housing stock, or by a movement of housing releases that does not lead to a dwelling being occupied within the study area. The model takes into account the mobilities specific to each household category, in terms – for example – of social position, age, size or the dwelling occupied. It provides information on trends in the population structures of the different housing types brought about by vacancies-reoccupancies. The article begins by describing the model’s theoretical foundations (filtering process and housing vacancy chains) and conception. It then goes on to present, through the example of the city of Lille (Nord, France), a method of data classification that allows comparative analysis suited to the application of the ASHA model. Next, it illustrates the model’s analytical scope and the way it can be used to understand the organisation of a housing system. Finally, drawing on a number of examples (family accommodation, gentrification, housing programmes), it sets out the model’s operational utility and the way local managers can use it to understand and anticipate the impact of housing policy on the urban population.