Thirty-nine subjects listened to two series of repetitive tones, the second series of which was of higher or lower pitch and of higher intensity. Before the first series all subjects were given conventional instructions to relax and adopt a passive attitude to the tones. Before the second tone series one of three instructions was given: to maintain a relaxed and passive attitude, to inhibit responding, to count the stimuli silently. The three groups showed roughly equivalent rates of habituation to the first tones series but the group instructed to count the tones showed slower habituation than the groups instructed to remain indifferent or to inhibit responses. A recommendation to clarify task demands in habituation experiments, by informing subjects that the aim was to see how quickly subjects stopped responding, showed no advantages over conventional instructions.