If the favored hierarchical cosmological model is correct, then the Milky Way system should have accreted ~100-200 luminous satellite galaxies in the past \~12 Gyr. We model this process using a hybrid semi-analytic plus N-body approach which distinguishes explicitly between the evolution of light and dark matter in accreted satellites. This distinction is essential to our ability to produce a realistic stellar halo, with mass and density profile much like that of our own Galaxy, and a surviving satellite population that matches the observed number counts and structural parameter distributions of the satellite galaxies of the Milky Way. Our model stellar halos have density profiles which typically drop off with radius faster than those of the dark matter. They are assembled from the inside out, with the majority of mass (~80%) coming from the \~15 most massive accretion events. The satellites that contribute to the stellar halo have median accretion times of ~9 Gyr in the past, while surviving satellite systems have median accretion times of ~5 Gyr in the past. This implies that stars associated with the inner halo should be quite different chemically from stars in surviving satellites and also from stars in the outer halo or those liberated in recent disruption events. We briefly discuss the expected spatial structure and phase space structure for halos formed in this manner. Searches for this type of structure offer a direct test of whether cosmology is indeed hierarchical on small scales.

22 pages, 16 figures, submitted to ApJ