The forming process for heterogeneous catalyst pellets impacts on reactor performance. In order to intelligently optimise the product formulation and manufacturing procedure, the impact of fabrication options on the resultant pore structure and mass transfer properties must be fully understood. In this work, the more direct information possible from multi-modal imaging methods has been combined with the more statistically-representative, multi-scale data from porosimetry, including the rarely-used gas overcondensation technique, to characterise batches of methanol synthesis catalyst tablets made using different feed types. Despite the hierarchies of complexity of the porous pellets revealed by the computerised X-ray tomography and FIB-SEM images, the impact on mass transfer of controlled modifications to the void space, achieved through mercury porosimetry, could be modelled using a relatively simple random pore-bond network. The characteristic parameters for the model were obtained from a percolation theory-based analysis of the overcondensation data. A quantitative relationship was thereby obtained between the structural information contained within the overcondensation isotherms and the rate of mass uptake of gas into the porous pellets. This revealed the differential importance to mass transfer of particular sets of pores, associated with certain pellet structural features, and the impact on tortuosity of pellet fabrication parameters such as particle feed size.