Debate exists as to the proper site for initiating antiarrhythmic therapy for supraventricular tachyarrhythmias and other benign forms of ectopy: inpatient versus outpatient. Rapid detection of efficacy and adverse effects, with immediate correction of the latter, favors the inpatient site. Convenience and, under most circumstances, lower cost favor the outpatient site. Circumstances under which adverse event rates, including proarrhythmia, are expectedly low, would favor outpatient initiation. So would the use of an agent whose elimination half-life is so long as to render in-hospital monitoring to steady state highly impractical. Accordingly, outpatient initiation would be suitable for patients without structural heart disease receiving class IC drugs, patients with low risk for torsades de pointes receiving selected class III agents, in whom data in the literature are supportive (as has occurred with sotalol and azimilide), and patients who are to receive amiodarone. Transtelephonic electrocardiographic monitoring can be used to facilitate assessment in the outpatient setting. Inpatient initiation should be considered for patients with underlying sinus node or atrioventricular conduction disturbances, for patients with significant structural heart disease, for patients receiving a drug whose proarrhythmia may be idiosyncratic (eg,quinidine), and for patients who are to begin an antiarrhythmic drug while in a supraventricular tachyarrhythmia in whom sinus rhythm has not previously been seen. The relative costs and benefits of the approach chosen will be a reflection of the probability that a drug with a chosen mechanism will cause an adverse outcome in a patient with a specific clinical substrate during the period chosen for monitoring.