The quinidine-like effects of some antidepressant drugs (particularly tricyclic antidepressants) and many antipsychotic drugs (particularly the phenothiazines) confound treatment of psychosis and depression in patients with major mental illness. This is especially true among elderly patients with existing risk factors for corrected QT (QTc) interval prolongation. We used PubMed, previously reported review articles and the extensive personal files of the authors to identify cases of subjects aged>or=60 years who developed QTc interval prolongation, polymorphic ventricular tachycardia (PVT)/torsade de pointes (TdP) and/or sudden cardiac death while taking antipsychotic or antidepressant drugs or a combination of these medications. We identified 37 patients who had taken, in total, 46 antipsychotic or antidepressant drugs. Our most striking finding was that almost four-fifths of our cases involved women. When the 14 critically ill subjects receiving haloperidol intravenously were excluded, 91.3% of our subjects were women. Almost three-quarters of our study subjects had cardiovascular disease. Intravenous administration of haloperidol in the critically ill and profoundly agitated elderly warrants particular comment. Of the 14 subjects in this category identified, six were men and eight were women. In 13 cases, the drug dose far exceeded the 2 mg necessary to produce an antipsychotic effect. These clinicians were using an agent to achieve sedation that usually requires very high doses in the critically ill and profoundly agitated elderly to achieve this effect. Inclusion criteria for our literature review required antipsychotic and/or antidepressant drug-induced QTc interval prolongation. Even so, our finding that 31 of our 37 subjects developed PVT is sobering. However, the reader should not conclude that drug-induced QTc interval prolongation is highly predictive of PVT or its TdP subtype. All of our study subjects had at least two risk factors for TdP, with age and sex being the most common. We included the rare case of a patient with congenital long QT syndrome who developed further lengthening of the QTc interval and TdP when prescribed an antidepressant drug well known to produce QTc interval prolongation. We conclude with recommendations for clinicians not expert in the specialty of cardiology to deal with the many questions raised in this review. Specifically, such clinicians treating elderly patients with antipsychotic and antidepressant drugs that may prolong the QTc interval should aggressively obtain a baseline ECG for elderly female patients with additional risk factors such as personal or family history of pre-syncope or syncope, electrolyte disturbances or cardiovascular disease. Elderly male patients are also subject to QTc interval prolongation when such risk factors are present. It is important that the clinicians themselves inspect ECGs. If the QT interval is more than half the RR interval, QTc interval prolongation is likely to be present. In such cases, a cardiology colleague interested in QTc interval issues and TdP should be asked to review the ECG. Finally, nothing in our recommendations replaces meticulous attention to US FDA guidelines in the package insert of each drug.