Overtly hypertriglyceridemic patients with non-insulin-dependent diabetes mellitus were given a control diet containing 120 g of sucrose and 50 percent carbohydrate, and later randomly assigned to receive isocaloric high- (220 g), intermediate- (120 g), or low- (less than 3 g) sucrose/carbohydrate diets for four weeks. The low-sucrose diet group demonstrated a modest but significant decrease in mean fasting serum glucose level in the first week only, although this change was no different from the other two dietary groups and was not sustained. All groups had little change in late postprandial serum glucose levels from control values, and no significant alterations in 24-hour glycosuria. The high-sucrose diet group demonstrated a significant increase in fasting serum triglyceride levels by the second week of the study, whereas the intermediate- and low-sucrose diet groups showed a decrease in mean fasting triglyceride levels. In contrast, the low-sucrose diet group's late postprandial serum triglyceride levels increased by the fourth week, whereas levels fell in the high-sucrose diet group. Mean fasting serum cholesterol concentrations decreased from control values in the high-sucrose diet group. Thus, although very high sucrose and carbohydrate consumption is clearly deleterious to fasting tryglyceride levels in non-insulin-dependent diabetes mellitus with preexisting hypertriglyceridemia, it appears that low dietary sucrose and carbohydrate proportions do not further improve preprandial glycemia and glycosuria and may adversely affect late postprandial serum triglyceride concentration. This study suggests that isocaloric sucrose and carbohydrate restriction below usual daily levels (120 g per day) offers no consistent benefit in glycemia or lipid control in overt type II diabetes.