The main objective of this paper is to test the capability of the Maxwell-Stefan theory for predicting the diffusion behaviour of multicomponent mixtures within micropores on the basis of the diffusion behaviour of single components. Diffusion within micropores involves movement of sorbed species. It is an activated process. The proper driving force is the gradient of the surface chemical potential. In the Maxwell-Stefan theory for micropore diffusion, the adsorption sites on the surface are viewed as pseudo-species, analogous to craters on the surface of the moon. The surface coverage has a significant influence on the mass transfer fluxes. The Maxwell-Stefan theory yields alternative approaches to micropore diffusion as special cases and is consistent with the theory of irreversible thermodynamics. The Maxwell-Stefan diffusivity for signal-component diffusion D is usually referred to as the "corrected" diffusivity in the literature. For binary diffusion, the Maxwell-Stefan equations, when combined with Langmuir adsorption equilibrium, coincide with the formulation derived by Habgood in 1958. To test the predictive capability of the Maxwell-Stefan theory we performed breakthrough experiments with single components and mixtures containing methane, carbon dioxide, propane and propene. These experiments were carried out in a packed bed of microporous activated carbon and with carbon molecular sieves. It is clearly demonstrated that the mixture behaviour can be predicted by the Maxwell-Stefan theory extremely well under a wide range of conditions : co-adsorption, co-desorption and counter-sorption. A model in which the matrix of Fick diffusivities is assumed to be constant is shown to be less successful in this regard.