The cytoskeleton is a complex, chemical heterogeneous network of semiflexible protein filaments, cross-linking proteins, and molecular motors that control the mechanics of some eukaryotic cells. Investigations of the mechanics of simplified biopolymer networks have shown that these materials differ substantially from better understood polymer gels in at least two< respects: (i) they show large deviations from the predictions of continuum elasticity in sparsely cross-linked networks but undergo a non-affine to affine cross-over in denser networks and (ii) when endogenous molecular motors are present to drive the network out of equilibrium, the elastic moduli of the nonequilibrium network increase by more than one hundred fold.In this talk we report analytic calculations and numerical simulations of equilibrium and non-equilibrium networks. Previous theoretical studies of the non-affine to affine transition in cytoskeletal networks have been confined to statistically isotropic random networks of monodisperse filaments. Biologically relevant and experimentally realizable networks are highly polydisperse and are frequently comprised of filaments with a preferred orientation. We examine, via numerical simulation, the individual effects of uniaxial order, filament polydispersity, and motor activity on the non-affine to affine transition. Finally, we demonstrate analytically how one can use the correlated motion of pairs of tracer particles embedded in the network (i.e. two-particle microrheology) to experimentally determine the density of active motors in in vitro networks and in living cells.