To evaluate arrhythmogenicity in patients who receive a modified direct-current (DC) shock ablation (distal pair of electrodes connected in common as the cathode) or radiofrequency (RF) ablation of supraventricular tachycardia, a prospective study was performed with signal-averaged ECG, 24-hour Holter monitoring, electrophysiologic study (EPS) for ventricular tachycardia (VT), and treadmill exercise test. Sixty-nine consecutive patients with documented paroxysmal supraventricular tachycardia were included. Twenty-eight patients proved to have atrioventricular nodal reentrant tachycardia, and 41 patients had atrioventricular reciprocating tachycardia that involved accessory atrioventricular pathways. The first 34 patients received DC shock ablation and the other 35 patients received RF ablation. Signal-averaged ECG, Holter monitoring, and EPS for VT were performed before ablation, immediately after ablation, then 1 week, 2 weeks (Holter monitoring), 1 month (except EPS), and 3 months after ablation. Treadmill exercise testing was performed before ablation, and at 1 week and 3 months after ablation. The root mean square, low-amplitude signal and QRS duration of signal-averaged ECG disclosed no significant change after either DC or RF ablation up to 3 months. Late potential developed in only one patient in the DC shock group and it was considered to be innocuous because neither VT nor ventricular fibrillation was noted or induced. Increases in the number of ventricular premature contractions and in short-run VT were detected by Holter monitoring in the first week after either mode of ablation ( p < 0.001 for the DC shock group; p < 0.05 for the RF group), which were greater ( p < 0.05) and lasted longer in the DC shock group than in the RF group. DC shock also caused an increased number of atrial premature contractions in the first week after ablation ( p < 0.05). No sustained VT, nonsustained VT, or sudden death was noted in either group during the follow-up period (3 months to 1 year). Sustained VT or ventricular fibrillation was not induced by EPS in either group up to 3 months after ablation. No exercise-induced VT was observed in either group in the study. In conclusion, these ablation methods were found to be relatively safe in treating patients with supraventricular tachycardia, since only transient atrial and ventricular irritability were observed. However, longer observation is required to evaluate long-term safety.