A study was conducted to test the hypothesis that high-variability rehearsal more effectively facilitates word recall than does low-variability rehearsal. Third-grade and sixth-grade students were asked to memorize a list of 20 common words. A read-aloud rehearsal procedure was used so that rehearsal could be experimentally controlled. One word was presented every 5 seconds; on the same card, the rehearsal accompanying it was also presented. As each card was shown, the experimenter read the to-be-remembered word aloud, and then the child read the rehearsal aloud. The high-variability rehearsal protocols consisted of the word to be remembered plus the previous two items. The low-variability rehearsal protocols were designed so that any given item was likely to be rehearsed along with only two other items. Free recall was tested after each of four study/test trials, and a form of cued recall was measured at the end of the session. Results showed that sixth graders consistently remembered more words than did third graders. Both third and sixth graders recalled more words in the high-variability rehearsal condition. Also, results of the cued recall test suggested that more inter-item associations were developed through reading the high-variability rehearsal. (RH)