It is now well recognized that therapy with antiarrhythmic drugs can not only suppress cardiac arrhythmias, but also may increase their frequency or provoke new ones. Specific proarrhythmia syndromes, each with a distinct underlying mechanism and approach to therapy, have been described. The best-recognized examples are digitalis intoxication, proarrhythmia associated with sodium-channel block, and torsade de pointes occurring during QT-prolonging therapies. In the case of sodium-channel blockers, 2 forms of proarrhythmia are commonly recognized: slow atrial flutter with 1:1 atrioventricular conduction, and frequent ventricular tachycardia ([VT], most often found in patients with pre-existing VT reentrant circuits). In all cases, the best approach to therapy is to identify patients at risk (and thereby avoid therapy entirely), to recognize proarrhythmia when it occurs, to withdraw offending agent(s), and to use specific corrective therapies when available. Although most recognized episodes of proarrhythmia are thought to occur early in drug therapy, the increased mortality during chronic antiarrhythmic therapy demonstrated in large randomized trials suggests this phenomenon can also develop during long-term drug treatment. The recognition of proarrhythmia and the delineation of its underlying mechanisms should not only improve therapy with available drugs, but may also direct development of newer agents devoid of this potential.