The first study in this series [Houts PS, Bachrach R, Witmer JT, Tringali CA, Bucher JA, Localio RA. Patient Educ. Couns. 1998;35:83-8] found that recall of spoken medical instructions averaged 14% but that, when pictographs (drawings representing the instructions) accompanied the spoken instructions and were present during recall, 85% of medical instructions were remembered correctly. Those findings suggested that spoken instructions plus pictographs may be a way to give people with low literacy skills access to medical information that is normally available only in written form. However, there were three important limitations to that study: (1) the subjects were literate and perhaps literate people remember pictograph meanings better than people with low literacy skills; (2) only short term recall was tested and, for medical information to be useful clinically, it must be remembered for significant periods of time and (3) a maximum of 50 instructions were shown in pictographs, whereas managing complex illnesses may require remembering several hundred instructions. This study addresses those limitations by investigating 4-week recall of 236 medical instructions accompanied by pictographs by people with low literacy skills. Subjects were 21 adult clients of an inner city job training program who had less than fifth grade reading skills. Results showed 85% mean correct recall of pictograph meanings immediately after training (range from 63 to 99%) and 71% after 4 weeks (range from 33 to 94%). These results indicate that people with low literacy skills can, with the help of pictographs, recall large amounts of medical information for significant periods of time. The impact of pictographs on symptom management and patient quality of life remains to be studied.