Islet amyloid deposition in type 2 diabetes is associated with reduced beta-cell mass. Therefore, interventions aimed at reducing islet amyloid formation may help preserve beta-cell mass in type 2 diabetes. Rosiglitazone and metformin act by different mechanisms to improve insulin sensitivity and thereby reduce beta-cell secretory demand, resulting in decreased release of insulin and islet amyloid polypeptide (IAPP), the unique constituent of islet amyloid deposits. We hypothesized that this reduced beta-cell secretory demand would lead to reduced islet amyloid formation. Human IAPP (hIAPP) transgenic mice, a model of islet amyloid, were treated for 12 months with rosiglitazone (1.5 mg.kg(-1).day(-1), n = 19), metformin (1 g.kg(-1).day(-1), n = 18), or control (n = 17). At the end of the study, islet amyloid prevalence (percent islets containing amyloid) and severity (percent islet area occupied by amyloid), islet mass, beta-cell mass, and insulin release were determined. Islet amyloid prevalence (44 +/- 8, 13 +/- 4, and 11 +/- 3% for control, metformin-, and rosiglitazone-treated mice, respectively) and severity (9.2 +/- 3.0, 0.22 +/- 0.11, and 0.10 +/- 0.05% for control, metformin-, and rosiglitazone-treated mice, respectively) were markedly reduced with both rosiglitazone (P < 0.001 for both measures) and metformin treatment (P < 0.001 for both measures). Both treatments were associated with reduced insulin release assessed as the acute insulin response to intravenous glucose (2,189 +/- 857, 621 +/- 256, and 14 +/- 158 pmol/l for control, metformin-, and rosiglitazone-treated mice, respectively; P < 0.05 for metformin vs. control and P < 0.005 for rosiglitazone vs. control), consistent with reduced secretory demand. Similarly, islet mass (33.4 +/- 7.0, 16.6 +/- 3.6, and 12.2 +/- 2.1 mg for control, metformin-, and rosiglitazone-treated mice, respectively) was not different with metformin treatment (P = 0.06 vs. control) but was significantly lower with rosiglitazone treatment (P < 0.05 vs. control). When the decreased islet mass was accounted for, the islet amyloid-related decrease in beta-cell mass (percent beta-cell mass/islet mass) was ameliorated in both rosiglitazone- and metformin-treated animals (57.9 +/- 3.1, 64.7 +/- 1.4, and 66.1 +/- 1.6% for control, metformin-, and rosiglitazone-treated mice, respectively; P < 0.05 for metformin or rosiglitazone vs. control). In summary, rosiglitazone and metformin protect beta-cells from the deleterious effects of islet amyloid, and this effect may contribute to the ability of these treatments to alleviate the progressive loss of beta-cell mass and function in type 2 diabetes.