Modern treatment of cardiac arrhythmias is limited to pharmacotherapy, radiofrequency ablation, or implantable devices. Antiarrhythmic medications suppress arrhythmias, but their systemic effects are often poorly tolerated and their proarrhythmic tendencies increase mortality. Radiofrequency ablation can cure only a limited number of arrhythmias. Implantable devices can be curative for bradyarrhythmias and lifesaving for tachyarrhythmias, but require a lifetime commitment to repeated procedures, are a significant expense, and may lead to severe complications. One possibility is the use of gene therapy as an antiarrhythmic strategy. As an initial attempt to explore this option, we focused on genetic modification of the atrioventricular node. First, we developed an intracoronary perfusion model for gene delivery, building on our previous work in isolated cardiac myocytes and hearts perfused ex vivo. Using this method, we infected porcine hearts with Adbetagal (recombinant adenovirus expressing Escherichia coli beta-galactosidase) or with AdGi (adenovirus encoding the Galphai2 subunit). We hypothesized that excess Galphai2 would mimic the effects of beta-adreneric antagonists, in effect creating a localized beta-blockade. Galphai2 overexpression suppressed baseline atrioventricular conduction and slowed the heart rate during atrial fibrillation without producing complete heart block. In contrast, expression of the reporter gene beta-galactosidase had no electrophysiological effects. Our results demonstrate the feasibility of using myocardial gene transfer strategies to treat common arrhythmias.