In 1973, Jim Bardeen calculated that a black hole should cast a shadow more than twice the size of its event horizon, and imaging the immediate surroundings of a black hole has been a goal of astrophysics ever since. Ongoing very long baseline interferometry observations at millimeter wavelengths are on the cusp of realizing this goal by leveraging radio dishes across the world to form an Event Horizon Telescope. These observations will allow the precise study of accretion & outflows, may provide direct evidence for the existence of a black hole event horizon in nature, and could potentially test general relativity in the strong field regime. The two most promising targets are the Galactic center black hole candidate, Sagittarius A*, and M87, known for its kiloparsec-scale jet; and event horizon scale structure has been detected in both sources. Maximizing the science return on the observations requires detailed theoretical modeling of the accretion flows in both sources, and this problem is especially well suited to contemporary numerical simulations. I will discuss black hole images calculated from simulations and their comparison with current data. The models are in excellent agreement with observations, and already constrain the allowed parameter space for the black hole and its accretion flow. The black hole shadow may be accessible within the next 5-10 years, and its detection would constitute the first direct evidence for an event horizon.