Prevailing insulin resistance and the resultant hyperglycemia elicits a compensatory response from pancreatic islet beta cells (β-cells) that involves increases in β-cell function and β-cell mass. However, the sustained metabolic stress eventually leads to β-cell failure characterized by severe β-cell dysfunction and progressive loss of β-cell mass. Whereas, β-cell dysfunction is relatively well understood at the mechanistic level, the avenues leading to loss of β-cell mass are less clear with reduced proliferation, dedifferentiation, and apoptosis all potential mechanisms. Butler and colleagues documented increased β-cell apoptosis in pancreas from lean and obese human Type 2 diabetes (T2D) subjects, with no changes in rates of β-cell replication or neogenesis, strongly suggesting a role for apoptosis in β-cell failure. Here, we describe a permissive role for TGF-β/Smad3 in β-cell apoptosis. Human islets undergoing β-cell apoptosis release increased levels of TGF-β1 ligand and phosphorylation levels of TGF-β's chief transcription factor, Smad3, are increased in human T2D islets suggestive of an autocrine role for TGF-β/Smad3 signaling in β-cell apoptosis. Smad3 phosphorylation is similarly increased in diabetic mouse islets undergoing β-cell apoptosis. In mice, β-cell-specific activation of Smad3 promotes apoptosis and loss of β-cell mass in association with β-cell dysfunction, glucose intolerance, and diabetes. In contrast, inactive Smad3 protects from apoptosis and preserves β-cell mass while improving β-cell function and glucose tolerance. At the molecular level, Smad3 associates with Foxo1 to propagate TGF-β-dependent β-cell apoptosis. Indeed, genetic or pharmacologic inhibition of TGF-β/Smad3 signals or knocking down Foxo1 protects from β-cell apoptosis. These findings reveal the importance of TGF-β/Smad3 in promoting β-cell apoptosis and demonstrate the therapeutic potential of TGF-β/Smad3 antagonism to restore β-cell mass lost in diabetes.