Recently, attention has focused on a potential link between schizophrenia and diabetes, with speculation that this potential associationis stronger in patients who are prescribed atypical antipsychotics. Pharmacoepidemiological studies can help to evaluate this potential association. Source data on the incidence of diabetes in patients treated with antipsychotics is available in the FDA MedWatch database, prescription claims databases and other patient registries. These data indicate that antipsychotic drugs may increase the risk of developing diabetes and that there may be an interaction with age. However, current data are insufficient to accurately assess potential differences in the risk of diabetes between users of individual antipsychotic medications. In addition, antipsychotic treatment-emergent diabetes has several distinct features, notably relating to age of onset, gender ratio, rate of deterioration of glycaemic control, and independence from initial treatment emergent weight gain. Nonetheless, guidelines for the control of risk factors for diabetes developed for the general population appear to be applicable to patients with schizophrenia.