The objective of this study was to determine the effect of a snack bar containing uncooked cornstarch, equivalent to 1 1/2 starch/bread exchanges (bar 1), compared to a control bar (bar 2), on the incidence of nocturnal and morning hypoglycemia in subjects with diabetes. Adolescent campers and counselors with diabetes (n = 79) were randomly assigned to Group A (5 nights of snack bar 1 as the evening snack, followed by 5 nights of snack bar 2) or Group B (5 nights of snack bar 2 as the evening snack, followed by 5 nights of snack bar 1). Midnight and morning finger stick blood glucose levels were compared to determine the incidence of hypoglycemia (< 60 mg/dl) and hyperglycemia (> 250 mg/dl), and events were analyzed for the total cohort, Group A, and Group B and by glycated hemoglobin quartile to determine the effect of each bar on glycemia. For subjects with diabetes there was a significant decrease in the number of hypoglycemic episodes with bar 1 compared to bar 2 at midnight (total cohort and Groups A and B) and in the morning (total cohort and Group A). There was a significant decrease in the number of subjects to ever experience hypoglycemia with snack bar 1 compared to snack bar 2, a significantly lower incidence of hyperglycemic episodes at midnight with snack bar 1, and no difference in the incidence of hypoglycemia by glycated hemoglobin quartile. These data suggest that the snack bar containing uncooked cornstarch can diminish night time and morning hypoglycemia in subjects with diabetes, without causing hyperglycemia.