The purpose of this study was to determine the termination and acceleration rates for 1 to 6 attempts of antitachycardia pacing [ATP] delivered by ICD in order to terminate spontaneously occurring VTs. Twenty-four ICD recipients with active ATP programs, including a maximum of six ATP sequences and spontaneously occurring VTs during follow-up, were investigated. During a mean follow-up of 42 +/- 15 months (range, 17-63 months) 413 spontaneous VT episodes (17 +/- 14; range, 1-49 per patient) resulting in appropriate ATP delivery by the ICD occurred. ATP successfully terminated 328 episodes (80%) with a mean number of 1.6 +/- 1.1 pacing sequences. Eighty episodes (19%) were accelerated by ATP and 5 (1%) were unresponsive to ATP. The ATP success decreased until the third ATP sequence (59%-->31%-->24%), but increased again in the fourth to sixth attempt (46%-->46%-->29%). The acceleration rate increased from sequence one to sequence three (8%-->13%-->28%), but decreased significantly in further ATP attempts (19%-->0%-->0%). The mean time delays until redetection or termination after 4, 5, and 6 attempts of ATP were 22 +/- 5 seconds, 37 +/- 2 seconds, and 41 +/- 9 seconds, respectively. Nine patients (37%) used > or = 3 ATP attempts during follow-up and all of them had a therapeutic benefit from it. Five out of 13 VTs (38%) treated with > or = 4 attempts could ultimately be terminated by ATP. The results of this study demonstrate that the first ATP sequence is the most effective and that > 4 ATP attempts may be useful in a minority of patients. There seems to be a low risk of VT acceleration by the fourth to sixth ATP sequence. Because of the associated time delay, a high number of ATP attempts should only be programmed in patients with hemodynamically well-tolerated stable VTs.