This work studies pressure drops through several textile fabrics. A preliminary study of cloth characteristics, including scanningelectron micrographs, shows their speci"cities towards particular media. For 20 di!erent cloths, in terms of weave and raw material(rayon or activated carbon"bers), an experimental study is carried out using a pilot-unit, in order to measure air and water pressuredrops through one layer of each cloth. Fluid Reynolds numbers range from 0 up to 2500 for both#uids. This experimental studyshows the in#uence of speci"c parameters of cloths (like weave) on their dynamic behavior. Furthermore, the swelling phenomenon of"bers in water is considered. Goodings'model is set up for woven structures and it enables the fabric opening diameter to becalculated around 10lm. Experimental data are then modelled,"rstly using classical models set up for particular porous media(Ergun, Carman's dimensionless model, Comiti}Renaud), and then using a statistical tool, neural networks. These models are testedusing three di!erent de"nitions for the speci"c surface area, on the fabric, yarn, and opening scale, respectively. Whichever thede"nition used, they are not suitable to describe the#ow through woven structures. However, they enable the swelling phenomenon of"bers in water to be con"rmed, and the#ow into the fabric yarn to be located. The experimental study, coupled with these modellingresults, leads to the choice of input neurons in the neural network (#uid properties*k,o,Re*and fabric characteristics*thickness, density, number of openingsNo,Soand raw material) in order to predict pressure drops as the output neuron. Thestatistical results obtained with this architecture are satisfactory and a variable analysis carried out with connection weight valuesenables the in#uence of speci"c parameters of cloths (likeNo) on pressure drops to be quanti"ed.(2000 Elsevier Science Ltd. Allrights reserved