A delayed matching-to-sample (DMTS) task was used to test several hypotheses about sources of retroactive inhibition (RI) in pigeon short-term memory. After being trained on DMTS with colors and patterns, pigeons were tested for RI in Experiments 1 and 2 by presenting interpolated stimuli designed to test the influence of the dimensions of similarity, saliency, familiarity, complexity, and degree of illumination. Of these dimensions, the only one which proved to be a significant source of RI was degree of illumination. Experiments 3 and 4 demonstrated a direct relationship between degree of illumination emitted from either the center key or houselight and amount of RI. Possible mechanisms which could explain the strong influence of interpolated illumination upon retention in animals are discussed.