The concept of metabolic syndrome in psychiatry provides a united front for confronting a series of metabolic changes that are predictive of cardiovascular disease (CVD) and type 2 diabetes mellitus (T2DM), which are highly prevalent in severe mental disorders (SMDs), such as schizophrenia, bipolar disorders, and severe depression. This review attempts to answer the following questions: (1) Is there evidence of significantly increased risk of metabolic syndrome in SMDs? (2) How is this evidence explained by stress theory and functional polymorphism? (3) What role can psychopharmacology and psychosocial therapies play in minimizing the problem? We have done a historical review using related literature from Medline. Compared with the general population, metabolic syndrome is two to three times more common in SMDs. The evidence for this predates the era of antipsychotic drugs. Altered glucose metabolism and dyslipidemia seem to be integral to SMDs. However, major psychotropic drugs are associated with metabolic syndrome, because of their activity at the appetite-stimulating receptors. SMDs seem to trigger a pathogenic cycle that fuels metabolic syndrome. To explain these findings, a neural diathesis-stress model has been proposed. Furthermore, candidate genes associated with receptors for weight gain are implicated. Using metformin (≥750 mg/day) may significantly reduce metabolic risks, and the data support consideration of this intervention for psychiatric patients taking antipsychotics. The obstacles to the implementation of the available guidelines for monitoring metabolic effects and changing unhelpful lifestyles need to be overcome by making monitoring mandatory and integration of physical exercise into routine care. Drug development and genotyping for the risk factors are future solutions.