Over the past few years some manufacturing processes involving resin impregnation in dry preforms prior to cure, and more specifically resin infusion across the fabric thickness (the so-called Resin Infusion Processes) have been developed. In these increasingly used processes, the transverse permeability of the fabric controls both manufacturing cycles and dimensions, and consequently the mechanical properties of the final composites. A big effort has been realized to obtain reliable experimental techniques for measuring the transverse permeability. Since the distinction between saturated and effective permeabilities is still being discussed, and the flow-front is hardly defined in such small dimensions (of the order of 1 mm), our measurements on some so-called Non-Crimped New Concept (NC2) are carried out for saturated flow, i.e., the fiber network is completely impregnated right from the onset of the experiment. In our case , these measures rely on the measurements of the pressure drop induced by the flow of a controlled fluid across the fabric, easily converted into transverse permeability through Darcy's law. One of the main results of these measurements is that the low NC2 transverse permeability depends on the face of the fabric receiving the fluid. This original behavior may, in turn, be translated into manufacturing guidelines to achieve optimal processing configuration. An attempt to relate this differential to the stitching hole surface is proposed which appears to explain some minor phenomena. On the contrary, the 3D shape of the stitching hole reconstruction from polishings could yield some appropriate explanation.