Four experiments were conducted to test the notion that rehearsal in a verbal discrimination learning task is not confined to the feedback interval, i.e., displaced rehearsal. Specifically, it was hypothesized that during early learning trials rehearsal would be restricted to the feedback interval. However, during later trials, once the subject was beyond the point of guessing the correct item, rehearsal could also occur during the anticipation interval. In all four experiments the opportunity for rehearsal was manipulated by presenting factorial combinations of either short or long anticipation and feedback intervals. Contrary to the hypothesis, no consistent differences were noted across trial blocks. However, the evidence did indicate that a long anticipation interval combined with a short feedback interval facilitated learning relative to a short anticipation interval combined with a long feedback interval. A modified version of displaced rehearsal provides a reasonable explanation for this finding if it is assumed that correct recognition is not a prerequisite for displaced rehearsal.