Type 2 diabetes is characterized by insulin resistance and a gradual loss of pancreatic beta cell mass and function1,2. Currently, there are no therapies proven to prevent beta cell loss and some, namely insulin secretagogues, have been linked to accelerated beta cell failure, thereby limiting their use in type 2 diabetes3,4. The adipokine adipsin/complement factor D controls the alternative complement pathway and generation of complement component C3a, which acts to augment beta cell insulin secretion5. In contrast to other insulin secretagogues, we show that chronic replenishment of adipsin in diabetic db/db mice ameliorates hyperglycemia and increases insulin levels while preserving beta cells by blocking dedifferentiation and death. Mechanistically, we find that adipsin/C3a decreases the phosphatase Dusp26; forced expression of Dusp26 in beta cells decreases expression of core beta cell identity genes and sensitizes to cell death. In contrast, pharmacological inhibition of DUSP26 improves hyperglycemia in diabetic mice and protects human islet cells from cell death. Pertaining to human health, we show that higher concentrations of circulating adipsin are associated with a significantly lower risk of developing future diabetes among middle-aged adults after adjusting for body mass index (BMI). Collectively, these data suggest that adipsin/C3a and DUSP26-directed therapies may represent a novel approach to achieve beta cell health to treat and prevent type 2 diabetes.