A simple deterministic model is proposed to represent the basic aspects concerning the effects of different antiretroviral treatment schedules on HIV incidence and prevalence of affected populations. The model mimics current treatment guidelines applied in Brazil. However, the model does not intend to fit the data with any acceptable degree of accuracy since uncertainties on the values of the parameters and on the precise effect of the treatment put some limits on the practical implications of our model from which only orders of magnitude and some qualitative effects can be deduced. So, this paper intends to provide a conceptual and mechanistic understanding of the possible long-term effects of treatment on the dynamics of HIV transmission. According to the model, the effect of the treatment depends on the level of sexual activity of the subpopulations considered, being more pronounced on the subpopulations with the highest sexual activity levels. Also, inefficient treatment can be prejudicial depending on the level of sexual activity and on the capacity to provide adequate treatment coverages to the population affected. (c) 2006 IMACS. Published by Elsevier B.V. All rights reserved.