Inhibition of the cardiac Kv1.5 channel, the molecular base for the human cardiac ultrarapid delayed rectifier potassium current (I-Kur), is considered a new promising atrial selective antiarrhythmic concept since this channel is presumed to contribute to atrial but not ventricular repolarization in the human heart. In a previous study in pigs we found clear baseline differences in refractoriness between left and right atrium with shorter effective refractory periods (ERPs) of the left atrium associated with a high left atrial vulnerability for tachyarrhythmias. In this newly established model we compared atrial and ventricular effects of two novel I-Kur blockers, S9947 and S20951, with the I-Kr blockers dofetilide, azimilide, ibutilide and d,l-sotalol. In pentobarbital anesthetized pigs (n=45) we determined ERPs in the free walls of both atria with the S1-S2-stimulus method at three basic cycle lengths (BCL 240/300/400 ms) and QTc-intervals. The incidence of atrial tachyarrhythmias triggered by the S2-extrastimulus of the left atrium was evaluated (referred to as left atrial vulnerability). In contrast to I-Kr blockade, I-Kur blockade had no effect on the QT-interval, but prolonged the atrial ERR The I-Kur blockers were significantly stronger on left atrial ERP, I-Kr blockers on right atrial ERP (P<0.05 for all compounds tested). At 240 ms BCL the I-Kur blocker S20951, 3 mg/kg, prolonged left vs. right atrial ERP by 28+/-5 ms vs. 12+/-3 ms and S9947, 3 mg/kg, by 45+/-7 ms vs. 19+/-6 ms. By contrast the effect of dofetilide, 10 mug/kg, was stronger on the right than left atrium (47+/-6 ms vs. 25+/-2 ms), a profile also found with azimilide (5 mg/kg, 43+/-3 ms vs. 17+/-3 ms), ibutilide (15 mug/kg, 70+/-10 ms vs. 29+/-4 ms) and d,l-sotalol (1.5 mg/kg, 57+/-6 ms vs. 36+/-4 ms). The I-Kur blockers, S20951 and S9947, significantly decreased left atrial vulnerability (-82% and -100%, respectively, P<0.01) in contrast to the selective I-Kr blocker dofetilide (-14%; n.s.). In conclusion, I-Kur and I-Kr blockers showed substantial differences in their atrial and ventricular actions in pigs. I-Kr blockers were stronger on right atrial ERP, I-Kur blockers on left atrial ERP, suggesting interatrial differences in the expression of potassium channels. In contrast to selective I-Kr blockade, I-Kur blockade inhibited left atrial vulnerability and had no effect on the QT-interval. Thus, blockade of I-Kur seems to be a promising atrial selective antiarrhythmic concept.