Antiarrhythmic drug-induced ventricular tachyarrhythmias have traditionally been identified on the basis of prolongation of the QT interval and the appearance of undulating peaks of sequential QRS complexes and T waves that give the tachycardia a ''twisting'' aspect (so-called torsades de pointes). It is now understood, however, that a precise lengthening of the QT interval is not a necessary or sufficient criterion for the development of torsades de pointes associated with drug therapy. Indeed, not all agents that have been linked with torsades de pointes cause significant lengthening of the QT interval. Moreover, sufficiently accurate determination of QT interval duration to judge to prolongation is confounded by a number of difficulties, including considerable intraindividual variability. It is possible that the development of abnormal giant U waves may be involved in the genesis of these tachyarrhythmias. Such U waves are believed to represent the summation of early afterdepolarizations occurring throughout the myocardium. The prepause ventricular rate determines the amplitude of the U wave, and larger U waves give rise to longer and faster runs of ventricular tachycardia. These arrhythmias tend to be self-perpetuating in a bigeminal feedback pattern. At present, our knowledge base about the relation of the QT interval and torsades de pointes is grossly incomplete. More basic and clinical data are required before hard clinical judgements can be made with any precision.