This paper presents a microeconomic model of residential location that explores the emergence of a mixed belt where residents and farmers coexist beyond a city. The model is based on integrating urban economics with cellular automata in order to simulate equilibrium patterns in 2D and through time. Households commute to a CBD and enjoy neighbourhood externalities that are a function of both local residential density and farmland, or open space. They bid on the competitive land market and locate so as to maximize utility. Incremental population growth changes the neighbourhood and leads to rent adaptations. With appropriate parameter values a mixed belt may emerge between the urban and agricultural specialized areas. Settlements within this mixed area are more or less clustered or scattered depending on preferences and neighbourhood size.