I n this issue of the Journal of Cardiovascular Pharmacology, van Middendorp et al 1 compare 2 antiarrhythmic drugs used for the conversion of atrial fibrillation (AF) with respect to their effects on conduction and contractility in dog ventricles paced for conduction delay. The clinical background is pharmacological conversion of AF in heart failure patients, many of whom are now treated by cardiac resynchronization therapy with biventricular pacing. The experimental protocol is clear and straightforward: compared to the nonselective sodium channel blocker flecainide, the newer atrial-selective drug vernakalant should be superior with respect to hemodynamics "at least in theory." 1 The authors spent much attention on appropriate drug concentrations used for the conversion of AF in humans, and employed sophisticated mapping techniques to detect even small conduction delays. The result of the carefully conducted study is surprising and rather simple: both drugs delay conduction and contractility to a comparable extent. In addition and somewhat unexpectedly, biventricular pacing reversed the negative inotropic effects of both vernakalant and flecainide (only in the setting of induced left bundle branch block tested).The study presented by van Middendorp et al may have been initiated by the interest to learn more about the extent to which concomitant antiarrhythmic drug therapy affects the hemodynamic benefit of biventricular pacing. However, beyond this point, the present results give important insights into the recent concept of atrial-selective antiarrhythmic drug therapy. 2 This concept requires at least 3 conditions:1. Target channels expressed selectively in the atrium. 2. Compounds that block atrial-selective channel selectively. 3. Blocking of atrial-selective channels results in increased refractoriness.The first condition seems to be fulfilled because the ultrarapidly activating potassium current (I Kur ) is conducted by the channel protein K v1.5 that is predominantly expressed in the atria. I Kur is, therefore, one of the candidate targets to achieve atrial-selective antiarrhythmic drug therapy. Low concentrations of the experimental potassium channel blocker 4-AP can be used to block I Kur selectively over I to (a current that is expressed both in atrium and ventricle). 3 In analogy to a blockade of potassium channels in ventricles, one would expect I Kur -blockade to robustly prolong APD and increase refractoriness without promoting life-threatening arrhythmias in the ventricle. Accordingly, the concept of I Kur -blockade has gained a lot of interest and lead to the development of dozens of compounds (for extensive review see Ford and Milnes). 4 As expected, one such compound, Xention D0101, neither prolonged APD in human ventricular preparations nor the QT c interval in healthy subjects, 5 arguing for the absence of actions on the ventricle. Although Xention D0101 markedly slowed the early phase of repolarization, APD 90 was almost unaffected. Similar results were made before with 2 other I Kur blocke...