SD3212 is a new antiarrhythmic drug which has class I, III, and IV effects. The purpose of this study was to elucidate the electrophysiological effects of this compound on a rabbit atrial fibrillation model, and to test a hypothesis that atrial fibrillation threshold is a quantitative indicator of atrial vulnerability. Whole hearts were excised from rabbits, and the aortas cannulated to perfuse the coronary arteries. Atrial fibrillation was induced with a burst stimulation of 50 Hz for 1 s while 3 microM acetylcholine (ACh) was perfused. When the right atrial appendage was paced at 200-ms intervals, SD3212 prolonged interatrial conduction time: control 30 +/- 1.2 ms, ACh 33 +/- 1.4 ms, ACh + SD 1 microM 37 +/- 2.4 ms, ACh + SD 3 microM 52 +/- 8.1 ms. The drug also prolonged the effective refractory period: control 80 +/-3.0 ms, ACh 48 +/- 3.8 ms, ACh + SD 1 microM 65 +/- 4.7 ms, ACh + SD 3 microM 98 +/- 15 ms. The rate of induction of atrial fibrillation by rapid pacing was 26% in Tyrode's solution, 85% in the presence of ACh, and 38% in the presence of ACh + SD 1 microM. The atrial fibrillation threshold decreased from 8.6 +/- 0.8mA (control) to 2.5 +/- 0.7 mA in the presence of ACh. It increased again to 7.8 +/- 1.0 mA in the presence of SD3212 (1 microM). SD3212 prolonged both the conduction time and refractory period. A reversed use-dependency was not prominent. These features caused antifibrillatory effects. Thus, the atrial fibrillation threshold seems to be a good quantitative indicator of atrial vulnerability.