To learn more about the metabolic effects of dietary fructose and sucrose, 12 type I and 12 type II diabetic subjects were fed three isocaloric (or isoenergic) diets for eight days each according to a randomized, crossover design. The three diets provided, respectively, 21% of the energy as fructose, 23% of the energy as sucrose, and almost all carbohydrate energy as starch. The fructose diet resulted in significantly lower one- and two-hour postprandial plasma glucose levels, overall mean plasma glucose levels, and urinary glucose excretion in both type I and type II subjects than did the starch diet. There were no significant differences between the sucrose and starch diets in any of the measures of glycemic control in either subject group. The fructose and sucrose diets did not significantly increase serum triglyceride values when compared with the starch diet, but both increased postprandial serum lactate levels. We conclude that short-term replacement of other carbohydrate sources in the diabetic diet with fructose will improve glycemic control, whereas replacement with sucrose will not aggravate glycemic control.