Pictographs have been used in nonliterate societies to help people remember spoken instructions and, today, they could be used to help nonliterate people remember spoken medical instructions. This study tested the hypothesis that pictographs can improve recall of spoken medical instructions. Twenty-one junior college subjects listened to lists of 38 actions for managing fever and 50 actions for managing sore mouth. One of the action lists was accompanied by pictographs during both listening and recall while the other was not. Subjects did not see any written words during the intervention and, therefore, relied entirely on memory of what they heard. Mean correct recall was 85% with pictographs and 14% without (P < 0.0001) indicating that pictographs can enhance memory of spoken medical instruction--often to a very high level. For this technique to be clinically useful, further research is needed on how to achieve accurate recall of large amounts of medical information for long periods of time by nonliterate patients. By viewing illiteracy as a memory problem, the large body of research on learning and memory can be utilized in designing education materials for this group.