[en] A dynamic model aimed on the assessment of the long-term consequences of an accidental contamination of an urban environments has been developed. The model was designed to assess the radiation exposure, as a function of time, of the different kinds of people that uses the contaminated environment, the relative contribution of each exposure pathway and to simulate the application of countermeasures and its effects on the reduction of surfaces contamination and on the exposure of the individuals and of the population. The model is an empirical one, mainly based on environmental data gathered after the Chernobyl and Goiania accidents, and takes into account climatic and population habits characteristic of tropical areas. The model was applied here to a contamination with the radionuclide ¹³⁷ Cs but can be easily adapted to other nuclides by changes on parameter values. An analysis of the variabilities associated to the model outputs regarding population habits, different kinds of urban environment and parameters uncertainty has shown that the main source of uncertainty on model predictions is associated to a correct knowledge of population characteristics, its habits and used of the contaminated environment. (author)