Forty-five consecutive patients with polymorphous ventricular tachycardia (PVT) were studied. The arrhythmia proved to be of a drug-related cause in 27 and due to an electrolyte disorder in four patients. Coexistent cardiac diseases without metabolic or drug-related abnormalities included ischemic heart disease in three, cardiomyopathy in three, and mitral valve prolapse in two. PVT was exercise-induced in four and associated with bradyarrhythmias in two. A prolonged QT or corrected QT interval was inconsistently related to the occurrence of PVT. In patients in whom PVT was induced by certain type I drugs, other type I antiarrhythmic drugs were usually either ineffective or resulted in aggravation of arrhythmia. For the group as a whole, treatment with lidocaine resulted in inconsistent beneficial effects, while cardiac pacing was almost universally effective for those with drug-induced PVT, regardless of the length of the QT interval. Long-term amiodarone therapy proved safe and effective for 12 of the 24 patients with drug-induced PVT who required long-term therapy for their original arrhythmia. We conclude that identification of PVT is the key clinical issue and that the QT interval is not necessarily the prime abnormality nor the variable to be considered in predicting success of therapy. Temporary cardiac pacing appears to be very effective in the short-term management of these patients. Use of type I antiarrhythmic agents in patients with drug-induced PVT generally resulted in aggravation of arrhythmia. In contrast, long-term amiodarone therapy for control of the original arrhythmia appears to be a promising approach for those with PVT associated with type I agents.