The resident cells of bone tissue are the micromachines responsible for maintaining and adapting the tissue structure to meet needs associated with bone’s dynamic, physiologic function. Mechanical load induced extravascular fluid flow provides the mechanical and chemical signals that modulate bone cell activity. However, the mechanisms by which cell scale processes are translated to functional adaptation at the organ scale are not clear. Predictive multi-scale models provide a means to test virtually the effects of specific model parameters, increasing efficiency and speeding the discovery of mechanisms underlying functional adaptation. This chapter reviews top-down computational modeling approaches to predict the interplay between mechanical loading of bone, load-driven fluid flow, and associated augmentation of molecular transport within bone. As underscored in recent studies, typically applied idealizations in geometry, as well as spatial distribution (anisotropy) and material properties of cells and tissues, deteriorate the fidelity of extravascular flow predictions. For example, idealization of pericellular fluid space geometries causes orders of magnitude underprediction of stresses imparted by fluid drag on cell surfaces. New, bottom-up approaches will help to elucidate the mechanical and chemical signals comprising the mechanophysiological environment of bone at multiple length scales, which is key to understanding mechanotransduction and how cells adapt bone tissue in health and disease.