This study examined the hypothesis that the glucose component of food and not the total carbohydrate is the major determinant of the glycemic response in patients with insulin-dependent diabetes mellitus. Patients were given glucose alone, fructose alone, glucose + fructose, lactose, and glucose + fat + protein. Fructose given alone increased the blood glucose almost as much as a similar amount of glucose (78% of the glucose-alone area, p less than 0.05). However, the same amount of fructose given with glucose produced no greater glycemic response than did glucose alone (108%). Similarly, galactose contributed only slightly to the glycemic response when given as lactose (122%, p less than 0.01) whereas protein and fat had no additional glycemic effect (101%). To test the above hypothesis in natural foods, patients were fed an amount of bread (high glycemic index) or apple (low glycemic index) that contained 25 g glucose. Both challenges produced glycemic responses very similar to 25 g purified glucose.