Previous search experiments in this laboratory have been concerned with the marked differences in target detectability as a function of its location in the visual field-differences we have called a detectability gradient-when subjects were required to detect a vertically oriented "target" among a number of distractor items having different orientations. This gradient was characterized by a marked right visual field superiority as well as differences in the shape of the gradient in the two half fields. A scanning model was proposed to account for these robust phenomena. The present experiments, using reaction time methods with vertical, horizontal, and colored targets have revealed marked differences in both reaction time and detectability as a function of target location and have isolated some of the spatial nonuniformities in the visual system which influence the shape of these gradients. The results reported here have forced us to abandon the scanning hypothesis in favor of a model in which attention is efficiently guided to the target.