High surface area Ce1 − xZrxO2 powders (0 < x < 0.36) were annealed at 950 °C. The surface area variation versus time of calcination was followed by BET method for all the samples. When the experimental rates of the surface area decrease are plotted versus x, two extremes are observed: a minimum around 0.09 and a maximum around 0.19. Our previous model involving oxygen and cerium diffusion proposed to interpret the stabilizing effect of zirconium at low content in ceria cannot account for the present experimental rate variations. We propose that the assumption of the ideal behaviour of the solutions is no longer valid when x reaches values as high as 0.32 and 0.36. Thus, provided that the Ce1 − xZrxO2 phases are considered as strictly regular solutions and on the basis of the kinetic model for ceria, it can be shown that the present experimental rate can be interpreted