Evaluating fluid leakage rates through cracks in concrete structures is essential in many fields (nuclear safety, fluid storage structures, durability of structures subjected to environmental conditions). In previous works, the authors proposed experimental procedures to study the water (Rastiello et al., 2014, 2018) and air (Tailhan et al., 2018) leakage processes through cracking concrete samples under Brazilian loading conditions. In a recent work, Tailhan et al. (2020) proposed a critical analysis of their own air permeability experimental setup considering an idealized parallel-plate system equipped with additional pressure sensors inside the crack. They showed that calibrating air flow models based on this experimental setup (and, more generally, experimental setups using “large” pressurized vessels placed upstream/downstream of the cracked sample) requires the knowledge of the pressure distribution inside the crack and that simply knowing the pressures inside the upstream and downstream chambers can lead to a misinterpretation of experimental results. In the present paper, a similar experiment is performed on a so-called “realistic crack” (i.e., representative of a crack surface in a split concrete block) molded in room-temperature-vulcanizing silicone equipped with internal pressure sensors. The results obtained show that knowing the real pressures at the inlet and outlet sections of the crack is even more important than in the case of the parallel-plate channel. Indeed, in that case, additional non-linear effects are observed in the mass flow-rate vs. square pressure gradient relation due to the geometry of the crack. Moreover, the crack geometry strongly affects the pressure drops measures at the inlet and outlet crack sections. These results clearly illustrate that attention has to be paid to interpreting experimental results on real cracks in concrete in the absence of pressure measures inside the crack. To the best of the authors’ knowledge, no experimental setups of this kind exist.