Bread is a foam and the softness of breads is an important quality attribute. For many non-food foams, digital environments have been implemented which can be used to characterize the structure to optimize products for specific functionalities. For breads, structure based models that incorporate pore structure, which is comprised of the pore distribution and overall porosity, with solid matrix properties are lacking. For the first time, such structure based models have been stochastically, or randomly, generated for a range of breads using previously published X-ray microtomography data (Wang et al., 2011). The Weibull probability distribution has been used to characterize closed pore distributions, stochastic probability theory has been used to obtain representative porosities, and digital bread crumb models which incorporate pore structure have been created. Digital bread crumbs have been compared against real world samples via visual texture and small strain compression experiments. Digital bread crumbs exhibited similarities to real products in terms of cell wall thicknesses as seen from surface appearance. Results from digital compression' experiments using finite element analysis showed differences in Young's moduli between breads can be attributed to both differences in pore distribution as well as porosity. Crown Copyright (C) 2013 Published by Elsevier Ltd. All rights reserved.