Eighty-five diabetic patients who were proficient in English were studied to assess the impact of educational material of varying literacy levels on patient comprehension. Two samples of available diabetes foot care material of Grade 11 and 9 readability (measured by SMOG formula) and purposely written in-house material of Grade 6 readability were used. Patients were randomized to read information of either Grade 6 and Grade 11 or Grade 6 and Grade 9 readability. Socio-demographic data and reading habits were collected to allow for identification of literacy markers. The mean CLOZE score (a measure of comprehension) was better in patients who read the Grade 6 information than for both the Grade 11 and Grade 9 information (59.5 +/- 11.8, 46.8 +/- 22.0, 45.8 +/- 22.2 respectively, p < 0.001). When evaluated in terms of percent of patients that could independently understand the material, Grade 6 information outperformed the Grade 11 and Grade 9 information (60%, 19%, 21%, respectively, p < 0.001). For the Grade 11 and Grade 9 information, poorer comprehension was associated with a non-English speaking background, early school leaving age, infrequent reading habits, and preference for tabloids (p < 0.02). Comprehension when patients read the Grade 6 information was no longer dependent on two of the four identified literacy markers. We conclude that reducing literacy demands of health literature improves patients' comprehension. Attention to socio-demographic data and reading habits can assist educators in assessing patients' literacy status and ensuring patients are given literature of compatible readability.