Pulmonary fibrosis (PF) is an interstitial lung disease characterized by scar tissue formation in the lung parenchyma resulting from aberrant wound healing. Since collagen is the stiffest material in the lung, an increase in either its amount or its degree of cross-linking can elevate both regional and global lung stiffness. Elevated stiffness is not just the end-result of PF; it also drives disease progression via an aberrant mechanotransduction process, the conversion of mechanical cues to cellular signals, which leads to further deposition and/or cross-linking of collagen in a positive feedback loop. Computational models that link multiscale structure to function can advance the understanding of the clinical manifestations of PF. The biologic processes involved in PF signaling and progression can be modeled via systems of coupled differential equations. However, a disconnect exists between the microscale fibrotic process and the development of clinical abnormalities, which can be explained by the concept of percolation. The most complete models of PF to date combine physics-based network representations of the alveolar parenchyma with agent-based representations of fibrotic cell biology. These models offer an explanation for the emergence of characteristic tissue structure in PF such as subpleural honeycombing.