OBJECTIVE: To examine the effects of 4 weeks of subcutaneous administration of pramlintide, a synthetic analog of human amylin, on metabolic control in patients with type 2 diabetes using insulin. RESEARCH DESIGN AND METHODS: Serum fructosamine, HbA1c, and fasting plasma lipids were measured in 203 patients in a randomized double-blind placebo-controlled parallel-group multicenter trial using doses of 30 μg q.i.d., 60 μg t.i.d., and 60 μg q.i.d. RESULTS: Statistically significant reductions in serum fructosamine concentrations were observed in the pramlintide 30 μg q.i.d. group (17.5 ± 4.9 μmol/l, P = 0.029), the pramlintide 60 μg t.i.d. group (24.1 ± 4.9 μmol/l, P = 0.003), and the 60 μg q.i.d. group (22.6 ± 4.1 μmol/l, P = 0.001) compared with the placebo group (3.5 ± 3.8 μmol/l). There were also statistically significant shifts in the proportion of patients with an abnormal serum fructosamine concentration at baseline that normalized at week 4 within the pramlintide 60 μg t.i.d. group and the 60 μg q.i.d. group. Consistent with the fructosamine results, there were statistically significant reductions in HbA1c in the pramlintide 30 μg q.i.d. group (0.53 ± 0.07%, P = 0.0447), the pramlintide 60 μg t.i.d. group (0.58 ± 0.07%, P < 0.0217), and the pramlintide 60 μg q.i.d. group (0.51 ± 0.08%, P = 0.0242) compared with the placebo group (0.27 ± 0.08%). Total cholesterol concentrations were also statistically significantly reduced in both the pramlintide 60 μg t.i.d. group (8.4 mg/dl, P < 0.01) and 60 μg q.i.d. group (10.5 mg/dl, P < 0.01) compared with placebo (1.2 mg/dl). Body weight decreased in both of the pramlintide 60 μg groups, but the trend did not achieve statistical significance. The incidence of hypoglycemia was similar in all treatment groups. CONCLUSIONS: Reductions in serum fructosamine, plasma total and LDL cholesterol concentrations, and HbA1c support the hypothesis that pramlintide may improve metabolic control in patients with type 2 diabetes using insulin.