A multifaceted educational program was designed to maximize dietary adherence of twenty-three diabetic subjects in order to study two experimental diets for forty weeks. The results of the program showed that a high level of diet adherence was achieved in spite of the strict dietary guidelines. The success of the educational approach was also reflected in the high attendance rate (99 per cent of 270 subject visits) and the low voluntary attrition rate (4 per cent). Five aspects of the over-all design were felt to have been of major importance in gaining subject cooperation: (1) the small, grouporiented learning process; (2) frequent intervals of follow-up; (3) feedback on laboratory data; (4) individualization of diet prescriptions; and (5) family involvement. Since the majority of diabetics seen in medical clinics fail to follow their prescribed diets, the concepts and approach applied to this study may be applicable to the dietary management of diabetic patients. Such implications and specific recommendations are presented in this report.