Permanent pacemakers have been used for the treatment of tachycardias. Four studies have been performed to improve termination algorithms. Permanent extrastimulus pacing was assessed in 19 patients. During follow up of 13 to 36 (mean 27) months, 7 also required additional antiarrhythmic drugs. Pacing was effective in all, but there was one unexplained sudden death. Autodecremental (rate-increasing) atrial pacing was used in 20 patients with junctional arrhythmias. It was effective in all, especially with a burst duration of 5000 ms, and caused no acceleration of tachycardia. In contrast, constant rate overdrive pacing produced atrial flutter or fibrillation in 4 patients. Concertina pacing (up to 7 stimuli) was tried, and was effective, in 19 patients. In three patients, using one or two stimuli of short coupling intervals, atrial arrhythmias were induced. As the number of stimuli were increased, longer pacing cycle lengths became effective. The effect of reset has been suggested as a useful method of searching for the termination zone. Ventricular pacing reliably caused reset, but atrial pacing produced inconsistent results. These studies show that adaptive, limited burst pacing modes are most effective methods, but reset is not useful in the right atrium.