A comparative study, using a crossover design, was carried out on 260-well controlled diabetic out- patients to investigate the metabolic effects of the A (50% cal carbohydrate, 30% cal fat, and 20% cal protein) and the B-diets (Study I), since the adherence of diabetic patients to the A-diet is very poor, and uncontrolled diabetes ensues. In Indonesian rural areas traditional medicines are still widely preserved; onion and green beans represent two of the frequent ones. Being faced with such facts, two separate randomized crossover studies were performed on 20 cooperative diabetic outpatients each; paired t-test was used for statistical analysis. Half of the patients were assigned to the B-diet (68% cal carbohydrate, 20% cal fat, 12% cal protein) plus 3 X 20 g fresh onion per day (Study II), or plus 3 X 200 g green beans per day (Study III) in the first week, and the B-diet only in the second week; the other half was assigned the other way around. Well and poorly controlled diabetics were used as experimental patients of Study II and Study III, respectively. Study I demonstrated that the B-diet showed potent hypocholesterolemic effect (31.75 mg%, p less than 0.001), maintained diabetic state, and did not raise the fasting serum triglyceride level. Study II resulted in significant decrease in blood sugar level (4.37 mg%, p less than 0.05) in the onion treated group, however, no blood lipid levels-changes occurred during both diets.(ABSTRACT TRUNCATED AT 250 WORDS)