Insulin/sulphonylurea-treated diabetics attending a busy university diabetic clinic were studied to determine whether issuing drug information sheets and/or age influenced understanding and behaviour regarding their disease and its treatment, especially with respect to avoiding hypoglycaemia. Patients were each asked 10 basic questions (each correct answer scoring 1), stratified by age (20 were less than or equal to 45 years and 91 greater than 45 years). According to a single-blind randomised protocol, they were issued or not issued with drug information sheets (providing information to correctly answer all 10 questions). After 2-3 months, 107 (88 aged greater than 45 years) were retested and asked whether they recalled an information sheet, read it themselves or had it read to them. Whether or not patients received sheets, corresponding mean aggregate scores were very similar in both age groups and there was no correlation with age. Second test scores yielded clinically and statistically significant increments in both the sheet and no sheet groups, respective mean aggregate scores increasing from 4.48 to 5.80 and 5.14 to 6.27 (P less than 0.001). Among patients issued with sheets, 32 who recalled reading them achieved the greatest improvement in mean scores (4.53 to 6.16, P less than 0.001). Active interaction/communication (participation in first test, recall and reading of information sheet) had a favourable educational impact irrespective of age, but merely issuing drug information sheets had no benefit.