Epicardially implanted Dsotalol polyurethane composite matrices for preventing ischemic ventricular arrhythmias were studied in open chest dogs under general anesthesia. D-sotalol was combined with a polyureapolyurethane (3:7) in solvent-cast films, which were characterized in vitro for their drug release at 37°C at pH 7.4 (0.05 M K2HP04). D-sotalol in vitro release occurred rapidly in an initial burst phase, with roughly 20% released within the first five min, and 90% by 60 min. Thereafter, an exponentially decreasing release rate was observed with matrix depletion by live hours. In the animal studies, the left anterior descending coronary artery (LAD) was occluded for 10 min on an hourly basis for up to live occlusions. 10 min prior to the third LAD occlusion, either a D-sotalol matrix or a vehicle matrix (control) was placed on either the ischemic or nonischemic left ventricular epicardium. The study was then continued observing the effects of matrix placement on occlusions 3,4, and 5.200 mg D-SOtdO matrices, which delivered a net dose of 1.2 mg/kg, effectively inhibited ventricular arrhythmias only if placed on the left ventricular ischemic zone. Placement of 200 mg D-sotalol matrices in the nonischemic zone was ineffective for significantly reducing the occurrence of ventricular arrhythmias. Furthermore, D-sotalol controlled release matrices were ineffective for preventing ventricular fibrillation (VF) regardless of dose or placement site. 200 mg ischemic zone D-sotalol matrices resulted in plasma sotalol levels in regional coronary venous samples ranging from 3.5 pg/ml to 10.4 &ml. However, peripheral sotalol levels obtained simultaneously ranged from 0.23 pg/ml to 0.78 ,ug/ ml. It is concluded that epicardial D-sotalol controlled release matrices inhibited ischemic ventricular arrhythmias, but not VF, if placed in the left ventricular ischemic zone during repeated LAD occlusions.