Aims/hypothesis The Edmonton Protocol for islet transplantation has provided hope for type 1 diabetic patients. However, this protocol requires lifelong immunosuppression, specifically sirolimus, a cellular antiproliferate. The effect of sirolimus on human pancreatic ductal cells (HDCs) is not known. This may be important since HDCs are believed to be islet precursors. Since neonatal porcine islets (NPIs), which contain many ductal precursor cells, could be a potential clinical source of islets, we also tested the effects of sirolimus on this tissue. Methods HDCs (n=4), NPIs (n=9) and human islets (n=5) were cultured with and without sirolimus (20 ng/ml) for 6 days. Results HDCs and NPIs cultured with sirolimus showed a 50 and 28% decrease, respectively, in cell number relative to control (p<0.05). Control cultures expanded 1.65-and 2.44-fold relative to time 0. Decreases in cell number of sirolimus-treated HDCs were not due to apoptosis as measured by TUNEL staining. No functional effects on human islets or NPIs were observed following static incubation with high glucose. Treatment of syngeneically transplanted and naïve BALC/c mice with sirolimus resulted in altered OGTT profiles with prolonged elevation of hyperglycaemia and weight gain. There was no difference in graft and organ insulin content between treatment groups. Conclusions/interpretation Our results indicate that sirolimus decreases ductal cell numbers in culture and alters glucose-stimulated insulin secretion in vivo. The administration of sirolimus to islet transplant recipients is likely to impair graft function as a result of decreasing ductal neogenesis and induction of insulin resistance.