Although work on class III antiarrhythmics remains at an early stage, these agents still appear to possess greater efficacy and less proarrhythmia than conventional class I agents in those experimental arrhythmia models considered to be most representative of the clinical situation. Although prolongation of repolarization carries with its own tendency for pause-dependent arrhythmogenesis (i.e., torsade de pointes), available data suggest that this may be a function of nonspecificity in potassium channel block rather than a general characteristic of class III activity. The availability of new and more selective blockers of specific cardiac potassium channels under development as class III agents have already helped to clarify basic questions about the ionic mechanism of repolarization in the heart, and one hopes that a growing clinical data base will eventually determine the relative safety and efficacy of these agents in preventing symptomatic and life-threatening arrhythmias.