Eighty consecutive newly diagnosed non-insulin-dependent diabetic patients were randomly allocated into two groups to compare two different patient education regimens. One group received individual dietary instructions by a nurse and the other a short, written leaflet given by a doctor. The principal aim of the dietary instructions was weight reduction. A significant weight loss and improvement in the control of diabetes occurred in both groups, and these changes were similar in the two groups. At the end of one year's follow-up, however, only 25% of the patients were satisfactorily controlled (fasting blood glucose less than or equal to 6.0 mmol/l). The degree of weight loss correlated only weakly with the improvement in the metabolic control. The degree of obesity and insulin secretion capacity as measured at the beginning of the study did not predict the improvement of glycaemic control during the study. At the end of the study a significant improvement was observed in serum lipids of patients with good control (fasting blood glucose less than or equal to 6.0 mmol/l) or weight loss (greater than 5 kg). In conclusion, both brief, written and individual dietary instructions induced a significant weight loss as well as improved glucose and lipid metabolism in newly diagnosed non-insulin-dependent diabetic patients, but satisfactory metabolic control was achieved only in a minority of the patients.