This paper presents a method to establish a relationship between internal microstructure and the effective gas diffusion coefficient in fibrous materials via a mesoscopic modeling approach and, when possible and based on the analysis, to propose user-friendly formulas as functions of structural parameters for practical engineering applications. The entire numerical framework includes two main parts: a random generation-growth method to reconstruct the digital microstructures of fibrous materials based on experimental statistical information of the actual structure, and then a high-efficiency lattice Boltzmann algorithm for modeling the gas diffusion process through porous structures. The predictions are then validated by existing experimental data for both dry and saturated fibrous materials. Owing to the unique robustness of the developed modeling approaches, we are then able to conduct a parametric analysis, more detailed than ever, of the influences on the system effective diffusion coefficient in fibrous materials by such important parameters as structural anisotropy, system water content, microstructure morphology and the layering space in a laminated fibrous system. These results may improve our understanding of gas diffusion in fibrous materials, and this method may serve as a tool for easy estimation of effective diffusivity, leading to the optimal design of fibrous materials. (C) 2016 Elsevier Ltd. All rights reserved.