This paper is concerned with building up methodological guidelines for modelling urban land‐use change through Geographical Information Systems, remote sensing imagery and Bayesian probabilistic methods. A medium‐sized town in the west of São Paulo State, Bauru, was adopted as a case study. Its urban structure was converted into a 100 m×100 m resolution grid and transition probabilities were calculated for each grid cell by means of the ‘weights of evidence’ statistical method and upon the basis of the information related to the technical infrastructure and socio‐economic aspects of the town. The probabilities obtained from there fed a cellular automaton simulation model—DINAMICA—developed by the Centre for Remote Sensing of the Federal University of Minas Gerais (CSR‐UFMG), based on stochastic transition algorithms. Different simulation outputs for the case study town in the period 1979–1988 were generated, and statistical validation tests were then conducted for the best results, employing a multiple resolution fitting procedure.This modelling experiment revealed the plausibility of adopting Bayesian empirical methods based on the available knowledge of technical infrastructure and socio‐economic status to simulate urban land‐use change. It indicates their possible further applicability for generating forecasts of growth trends both for Brazilian cities and cities world‐wide.