Batty M. and Karmeshu (1983) A strategy for generating and testing models of migration and urban growth, Reg. Studies 17, 223–236. In this paper, it is argued that to identify the appropriate conditions under which the traditional gravity-type migration model should be tested, it is necessary to consider the wider demographic-economic framework in which such models play an essential part. The traditional model and attempts to make it more realistic in an ad hoc way are first sketched, and then the model is embedded in a dynamic framework based on nonlinear differential equations of the logistic kind describing the growth of population and employment. The role of the gravity model in coupling these equations is then developed. A theoretical analysis of the framework using simulation follows. The simplest possible two-region case involving urban and rural regions is developed and after preliminary simulations, a more realistic version involving urban core, suburb and rural hinterland is proposed. A number of variants of this model are tested using simulation and it is concluded that in devising an appropriate test for this type of migration model, attention should be switched from spatial to temporal variation in migration using highly polarized spatial situations such as city regions.