During the last years, a contribution of antipsychotic drugs in the increase of diabetes prevalence in schizophrenic population has been repetitively suggested. The debate focused mainly on the second-generation antipsychotics. The analysis of the scientific literature indicates however that this discussion is not recent and an increase of diabetes prevalence in schizophrenic populations was already described before the introduction of neuroleptics. Then, after the introduction of the first neuroleptics in the 1950s, an increase of diabetes prevalence was reported among treated patients and the same alarms occurred in the 1990s after the introduction of second-generation antipsychotics. These treatments were related to an increase of glucose tolerance impairment, type II diabetes and diabetic acidoketosis. Recent epidemiological studies have confirmed the increase prevalence of diabetes in schizophrenic patients, particularly in schizophrenic patients before any antipsychotic treatment. Among the suggested mechanisms, there are sedentary life (due to hospitalisation and sedative effects of neuroleptics), food imbalance, shared genetic factors for diabetes and schizophrenia. Moreover, the frequency of the metabolic syndrome is increased in schizophrenic populations. This syndrome associates blood glucose increase, lipid metabolism disorders and android obesity. This could explain--via an increase of the cortisol production--the increase of mortality due to cardiovascular diseases observed in schizoprhenic patients. Thus, it seems well established that schizophrenia is associated with an increased risk for diabetes. It is however more difficult to evaluate the role of antipsychotic treatment as a causative factor of diabetes. Indeed, there are many published case reports or diabetes or diabetic acidoketosis after an antipsychotic treatment, but the level of evidence in controlled trials is low. Many studies were performed on large databases, but were retrospective and subjected to many flaws: concomitant diseases not taken into account, diabetes status evaluated by drug consumption, unknown diabetes status before antipsychotic treatment, etc. In the few prospective studies performed, no significant differences between the atypical versus typical antipsychotics were evidenced for new cases of diabetes. Moreover, in general population, the glucose tolerance impairment is underdiagnosed and it is estimated that people with a glucose tolerance impairment have a 5-10% annual risk of type II diabetes. Thus, this concern has to be replaced among the world epidemic increase of diabetes and in a population of patients whose the disease itself and life style are risk factors for diabetes. Some studies have explored the pathophysiological mechanisms that could support a diabetogenic effect of antipsychotics. Although it does not seem to be a direct effect of antipsychotics on insulin secretion by pancreatic cells, body weight increase has been evidence for both typical and atypical antipsychotics. However, it remains unclear whether this weight increase is responsible for a visceral adiposity, which is a risk factor better fitted to the cardiovascular mortality tha the body weight itself. Other hypotheses involving an effect on the leptin, which regulates the appetite, have been proposed. In waiting of new prospective controlled studies, and without denying the impact of antipsychotics on the glucose and lipid metabolisms (on the weight increase, for example), it should be recognized that the benefit/risk ratio remains largely in favour of the treatment, particularly for the atypical antipsychotics, more effective and better tolerated at the neurological level than the conventional antipsychotics. One of the benefits of the mainly articles in professional media about this concern is to draw attention on the metabolism disorders in schizophrenic patients, which are important risk factor of their frequent cardiovascular surmortality whatever the causes. Consequently, it is advised to monitor glucose and lipid metabolisms of schizophrenic patients before and during their treatment (body weight, fast blood glucose, blood cholesterol and triglycerides). In conclusion, schizophrenic patients are a population with an increased metabolic risk, which is a cause of their increased mortality. Although these data are known since a long time ago, this population does not benefit from the same metabolic follow-up than the non-schizophrenic population. The debate on the possible relationship between diabetes and antipsychotics should be also taken as a helpful recall of the necessity to follow simple rules of prevention and monitoring in this at-risk population. This should make it possible to preserve the benefit of the antipsychotics, the contribution of which in the treatment of schizophrenia is not any more to demonstrate.