Spaced retrieval training uses a schedule of practice trials distributed according to a learner's performance. The authors compared spaced retrieval to four alternative schedules of practice to determine whether it is more effective than other schedules for people with dementia. Participants practiced (a) pill names or (b) nonverbal sequences. Spaced retrieval did not produce long-term retention more often than other schedules of practice on either task. Participants with higher scores on one portion of the Hopkins Verbal Learning Test-Revised were more likely to show long-term retention of pill names regardless of schedule of practice (Experiment 1).