We calculate the density of states of a binary Lennard-Jones glass using a recently proposed Monte Carlo algorithm. Unlike traditional molecular simulation approaches, the algorithm samples distinct configurations according to self-consistent estimates of the density of states, thereby giving rise to uniform internal-energy histograms. The method is applied to simulate the equilibrium, low-temperature thermodynamic properties of a widely studied glass former consisting of a binary mixture of Lennard-Jones particles. We show how a density-of-states algorithm can be combined with particle identity swaps and configurational bias techniques to study that system. Results are presented for the energy and entropy below the mode coupling temperature. (C) 2003 American Institute of Physics.