Twenty Ss estimated the brightness of two sets of gray-white metric patterns. Set I patterns had figure-ground response characteristics which varied with relative area, whereas Set II patterns remained constant though relative area changed. Maximum brightness contrast occurred when the gray and the white areas were equal with contrast decreased as one of the component areas came to dominate. In contradiction to previous attentional models of brightness contrast, this effect was found for both sets of patterns; however, Set I patterns decreased at a slower rate. The average log magnitude estimates for Set I, attentional patterns, were significantly greater than those for Set II, sensory patterns, only at the most extreme relative areas. It was suggested that relative area, or sensory effects, caused the decrease and that attention influenced the rate of decrease.