Tropical forests are known for their tree species diversity, that is difficult to take into account in models. Even if some authors do not renounce to model separately every species, the usual solution consists in defining groups of species, then adjusting a set of parameters for each group. Groups may be built from ecological characteristics of the species, thus providing so-called 'functional groups', but they may also take into consideration extraneous information such as commercial categories. Defining groups in relation to a model of forest dynamics is however un-easy, as two species may appear similar with respect to a biological function, and at the same time different with respect to another function. Crossing all causes of singularities among species then brings a diversity of groups that is comparable to the species diversity itself. In this study, we address this issue by allowing a species to move from one species group to another, depending on the biological process that is concerned. We developed this approach with a matrix model of forest dynamics, for a tropical rain forest in French Guiana, at Paracou, focusing on the methodological aspects. As in other matrix models, forest dynamics is split into three components: recruitment, growth, and mortality. Five groups of species were defined at Paracou from species characteristics, in a previous study; we re-analyzed these data to build 5 recruitment groups, 5 growth groups, and 5 mortality groups. One species is then characterized by its recruitment group, its growth group and its mortality group, thus yielding at total 5 x 5 x 5 = 125 possibilities, which brings a more realistic view of the floristic composition of the forest, as if each species had been modelled separately. The resulting matrix model however do not have more parameters than it would have with 5 global species groups: if r, g and m are the numbers of parameters required to model recruitment, growth and mortality for one species, then in