Over the past decades, theoretical modeling has become an indispensable component of research into the hemodynamics of microcirculation. Numerous studies rely on modeling to provide quantitative insights into the interacting biophysical mechanisms that govern microcirculatory flow. The mechanical deformation of hematocytes has been addressed by continuum and molecular-informed computational models based on a growing body of experimental information. Theoretical analyses of single-vessel flow and blood rheology have led to a range of modeling approaches. Until recently, computational constraints limited direct simulations of multi-particle flows involving deformation and/or aggregation, but recent studies have begun to address this challenge. Network-level analyses have provided insights into the biophysical principles underlying the design of the microcirculation. This approach has been used to complement available experimental data and to derive empirical models of microvascular blood rheology. Continued increases in computational performance applied to current modeling techniques will enable larger scale simulations. In order to exploit this opportunity, integration of diverse theoretical approaches within a multi-scale framework is needed.