Chicks were first imprinted by exposing them to a moving training stimulus, B or C, that was projected onto a screen at one end of an experimental cabinet. Subsequently, subjects selectively approached the stimulus to which they had been exposed. On the following day, the chicks were placed into a chilled experimental cabinet (15°C) and received trials on which two stimuli (A and B) were projected onto screens located at opposite ends of the cabinet. If the subject approached Stimulus A, a stream of warm air was delivered; if it approached B, the trial terminated, and no heat was presented. For subjects that had been imprinted with Stimulus B, those in Group B, Stimulus A was novel, and Stimulus B was familiar. For chicks that had been imprinted with Stimulus C, Group C, both A and B were novel. In two experiments, Group B acquired the discrimination more rapidly than did Group C. This observation, made using a novel training procedure, was taken to support the suggestion that imprinting results in a form of perceptual learning in which the familiar imprinting object has become more discriminable from other novel objects.