To determine factors affecting the success rate of rapid atrial pacing in converting atrial flutter to sinus rhythm in medically treated patients, we prospectively used this technique for 120 consecutive episodes in a total of 110 patients (94 male, 16 female, mean age 63 +/- 14 years). Structural heart disease was present in 77%, and all patients were receiving antiarrhythmic drugs at the time of the procedure. Atrial flutter type I was present in 92 of 110 patients (84%), and atrial flutter type II in 18 of 110 (16%). Primary success rate (return to sinus rhythm either immediately or after < 10 min of atrial fibrillation) was 70% (71/102) for flutter type I, and 6% (1/18) for flutter type II (p < 0.001). Delayed success (conversion to sinus rhythm in > 10 min but < 24 h) was observed in 15 additional episodes of flutter type I (15%) and in 1 additional episode of flutter type II (6%). The only clinical factors predicting primary success were (a) characteristics of flutter waves on the 12-lead surface electrocardiogram, (b) duration of flutter (primary success rate of 81% if flutter < 1 month vs. 57% if > 1 month, p < 0.05), and (c) flutter rate (primary success rate of 78% if < 260/min vs. 56% if > 260/min, p < 0.05). In 6/71 episodes of flutter type I (8%), prolonged sinus pauses or severe bradyarrhythmias occurred after conversion to sinus rhythm.(ABSTRACT TRUNCATED AT 250 WORDS)