This work uses a method for the stochastic reconstruction of catalyst layers (CLs) proposing a scaling method to determine effective transport properties in proton exchange membrane fuel cell (PEMFC). The algorithm that generates the numerical grid makes use of available information before and after manufacturing the CL. The structures so generated are characterized statistically by two-point correlation functions and by the resultant pore size distribution. As and example of this method, the continuity equation for charge transport is solved directly on the three-dimensional grid of finite control volumes (FCVs), to determine effective electrical and proton conductivities of different structures. The stochastic reconstruction and the electrical and proton conductivity of a 45 mu m side size cubic sample of a CL, represented by more than 3.3 x 10(12) FVCs were realized in a much shorter time compared with non-scaling methods. Variables studied in an example of CL structure were: (i) volume fraction of dispersed electrolyte, (ii) total CL porosity and (iii) pore size distribution. Results for the conduction efficiency for this example are also presented. (C) 2010 Elsevier B.V. All rights reserved.