The Cold Dark Matter paradigm predicts that dark matter halos are abound with substructure. A detailed characterization of dark matter subhalos is important for understanding the evolution of satellite galaxies, for modeling time-delays and flux-ratio anomalies in gravitational lensing, for boosting the dark matter annihilation signal and for the survivability of fragile structures such as streams and disks. I discuss the abundance and evolution of dark matter subhalos based on a new analytical model, use the model to revisit the Too-Big-To-Fail problem, demonstrate important shortcomings in the studies of substructure based on numerical simulations, and present a comprehensive analysis of the segregation of dark matter subhalos within their host halos.
