Treatment with the atypical antipsychotics olanzapine and clozapine has been associated with an increased risk for deterioration of glucose homeostasis, leading to hyperglycemia, ketoacidosis, and diabetes, in some cases independent of weight gain. Because these events may be a consequence of their ability to directly alter insulin secretion from pancreatic -cells, we determined the effects of several antipsychotics on cholinergic-and glucose-stimulated insulin secretion from isolated rat islets. At concentrations encompassing therapeutically relevant levels, olanzapine and clozapine reduced insulin secretion stimulated by 10 mol/l carbachol plus 7 mmol/l glucose. This inhibition of insulin secretion was paralleled by significant reductions in carbachol-potentiated inositol phosphate accumulation. In contrast, risperidone or ziprasidone had no adverse effect on cholinergic-induced insulin secretion or inositol phosphate accumulation. None of the compounds tested impaired the islet secretory responses to 8 mmol/l glucose alone. Finally, in vitro binding and functional data show that olanzapine and clozapine (unlike risperidone, ziprasidone, and haloperidol) are potent muscarinic M 3 antagonists. These findings demonstrate that low concentrations of olanzapine and clozapine can markedly and selectively impair cholinergicstimulated insulin secretion by blocking muscarinic M 3 receptors, which could be one of the contributing factors to their higher risk for producing hyperglycemia and diabetes in humans. Diabetes 54:1552-1558, 2005 R ecent reviews of clinical databases have revealed that olanzapine and clozapine carry a higher risk for producing hyperglycemia, ketoacidosis, and new-onset type 2 diabetes than other second-generation antipsychotics (SGAs) or haloperidol, a first-generation antipsychotic (1-6). The use of olanzapine and clozapine is often associated with notable weight gain and dyslipidemia, which are known risk factors in the development of diabetes. However, several reports have described cases of hyperglycemia following olanzapine and clozapine treatment that were not associated with weight gain (7,8). Furthermore, cases exist where switching to other SGAs, such as ziprasidone or risperidone, resulted in the reversal of olanzapine-or clozapine-associated hyperglycemia, suggesting that fundamental differences exist among the SGAs (9 -11).The mechanisms responsible for the increased diabetes risk of olanzapine and clozapine are not known, but in contrast to other SGAs, both compounds are potent muscarinic receptor antagonists (12). This led us to consider the possibility that disruption of the cholinergic processes regulating insulin secretion is one of the underlying mechanisms for impaired glucose regulation. Therefore, we investigated the effects of several antipsychotics on cholinergic-stimulated insulin secretion and the activation of phospholipase C using isolated rat pancreatic islets. Since the cholinergic activation of insulin release is mediated through muscarinic M 3 receptors on -cells ...