A general methodology was presented for the optimization of the macropore network in porous catalysts with given intrinsic kinetics and given nanostructure. Macropores were introduced to reduce diffusion limitations. Macroporosity, macropore size (d), and the thickness of the nanoporous, catalytically active macropore walls (w) influenced the overall product distribution. The optimization was based on a reduced gradient method in combination with a multigrid method to solve the set of discretized partial differential equations representing diffusion and reaction in the hierarchically structured porous material, containing nanoporous catalyst and macropores. The conversion of a single reaction was maximized if w was sufficiently small, and d was constant throughout. This optimum corresponded to a situation where diffusion limitations inside the nanoporous walls of thickness ware avoided, so that the diffusion resistance is limited to the macropores only. In situations with multiple reactions, the optimization methodology might be used to obtain the catalyst structure that maximizes selectivity toward a particular product. This methodology was applied to the important problem of autothermal reforming of natural gas, for the on-board production of hydrogen gas for fuel cells. This is an abstract of a paper presented at the 8th World Congress of Chemical Engineering (Montreal, Quebec, Canada 8/23-27/2009).