Patients with unilateral frontal- or temporal-lobe lesions and normal control subjects studied multiple arrays of pictures and were tested for recall of the locations of the pictures. One condition consisted of three trials of the same pictures in different spatial arrangements, recall being tested immediately after each presentation. In a second condition (using different stimuli), the subject was given two trials with one set of pictures, but a new set of pictures was viewed on the third trial. All groups showed a build-up of proactive interference across trials using the same pictures, and a release of proactive interference when they studied new pictures. Patients with frontal-lobe lesions were more susceptible to proactive interference than were the other groups.