Pramlintide, a synthetic analog of amylin, improves postprandial hyperglycemia. We compared subcutaneous (s.c.) pramlintide injection with square wave pramlintide infusion in adolescents with type 1 diabetes (T1DM). Eight subjects with T1DM underwent two randomized studies. Subcutaneous pramlintide (dose = 5 microg/unit of insulin) bolus, was given one time and another time, the same dose was given as a 120-min s.c. infusion. Insulin dose was constant between studies. Gastric emptying was assessed with oral acetaminophen and [l-13C]glucose in meal. Plasma glucagon, pramlintide, and insulin concentrations were measured. Insulin concentrations (p < 0.99) between pramlintide injection versus infusion were similar; however, glucose concentrations were different (p < 0.0001), with the absence of hypoglycemia during pramlintide infusion [AUC (0-120 min) -0.07 +/- 0.2 versus 1.05 +/- 0.24 mg * h/dL (p < 0.0088)]. Insulin-only administration resulted in postprandial hyperglycemia and late postprandial hypoglycemia (p < 0.0001). Two subjects experienced hypoglycemia with pramlintide injection. Pramlintide bolus caused pronounced glucagon suppression (p < 0.0003) and delayed gastric emptying as ([13CO2] p < 0.0003 and acetaminophen p < 0.01) compared with infusion. We conclude that pramlintide bolus may result in an increase in risk of immediate postprandial hypoglycemia. Further modifications in pramlintide delivery are indicated before it can be safely used in children.