Normal participants (n=5) having no experience with antipsychotic drugs and medicated participants (n=5) with clinical experience with chronic low doses of haloperidol (3-10 mg/day for 2-4 months) in the treatment of neuroses were evaluated for the effects of inter-trial interval (ITI) feedback on a discrete-trials peak-interval timing procedure. Feedback was presented during the ITI in the form of a histogram showing the distribution of the responses participants made on the previous trial plotted on a relative time scale. As feedback concerning the accuracy and precision of a reproduced duration (e.g., 7- and 14-s visual signals) became more remote in time, reproduced intervals gradually lengthened in duration. This rightward horizontal shift in peak time increased as a function of the probability of feedback and was enhanced by chronic treatment with haloperidol in a manner that was proportional to the duration of the signal. Our data suggest a gradual change in the underlying representation of the signal duration as a function of the remoteness of ITI feedback that is dependent upon both changes in working memory and the speed of the internal clock used to time durations in the seconds-to-minutes range.