This work investigates the capability of a fully predictive computational fluid dynamics method to catch the main features of binary segregating fluidized beds of particles differing in size and density. In particular, the numerical solution of a multiphase fluid-dynamic model based on an Eulerian description of both the continuous and the two dispersed phases and the kinetic theory of granular flows for modelling the inter-particle interactions is considered. The results obtained under a number of different operating conditions are compared with experimental data reported in the literature and the effect of grid size, time step and discretization schemes for the convective terms on axial segregation is also discussed. The capability of the model to correctly predict the minimum fluid velocity at which the mixture becomes steadily fluidized and fully mixed is also tested. Our calculations demonstrate that firm conclusions on the CFD model reliability in the prediction of the segregation dynamics in fluidized beds can be drawn only after careful verification of numerical uncertainties at the operating conditions under investigation. (C) 2011 Elsevier Ltd. All rights reserved.