AbstractCombining digital imaging, physical models and laboratory measurements is a step further towards a better understanding of the complex relationships between the soil pore system and soil functions. Eight natural 100‐cm3 soil cores were sampled in a cultivated Stagnic Luvisol from the topsoil and subsoil, which we assumed had contrasting pore systems. Artificial 100‐cm3 cores were produced from plastic or from autoclaved aerated concrete (AAC). Eight vertical holes of each diameter (1.5 and 3 mm) were drilled for the plastic cylinder and for one of the two AAC cylinders. All natural and artificial cores were scanned in an X‐ray CT scanner and printed in 3D. Effective air‐filled porosity, true Darcian air permeability, apparent air permeability at a pressure gradient of 5 hPa and oxygen diffusion were measured on all cores. The active pore system characteristics differed between topsoil (sponge‐like, network of macropores of similar size) and subsoil (dominated by large vertical macropores). Active soil pore characteristics measured on a simplified pore network, that is, from artificial and printed soil cores, supported the fundamental differences in air transport by convection and diffusion observed between top‐ and subsoil. The results confirm the suitability of using the conceptual model that partitions the pore system into arterial, marginal and remote pores to describe effects of soil structure on gas transport. This study showed the high potential of using 3D‐printed soil cores to reconstruct the soil macropore network for a better understanding of soil pore functions.