The efficiency of different technological processes where nanoporous carbons are used depends on their storage capacity and an appropriate gas transport. Different experimental and theoretical works that relate storage to global structural parameters of the solid such as specific surface area (SSA) or total pore volume (V-t) can be found in literature. The structure-transport relationships have been less studied. The combined use of the truncated pore network model (TPNM) and the kinetic Monte Carlo (KMC) method is proposed in this work to find hydrogen and methane effective self-diffusivities (D-eff) and to delve into those relationships. It was found that for Knudsen and free molecular diffusion in the simulated materials, the D-eff/V-t vs. SSA graphic nearly follows a power law. The KMC/TPNM approach was also used to predict the self-diffusion coefficients of hydrogen in Vulcan XC-72 and of methane in a carbon aerogel. The obtained values are within the expected range. KCM/TPNM is computationally fast and it allows a study of the diffusion synchronously and globally in the network, avoiding thus its fractionation in single pores and the use of just one geometric model to describe the porous spaces. (C) 2011 Elsevier Ltd. All rights reserved.