Background: Spinal cord stimulation (SCS) reduces the incidence of ventricular tachyarrhythmias in experimental models. This study investigated the effects of long-term SCS on ventricular function in a postinfarction canine heart failure model., Methods and Results: In stage 1, dogs underwent implantable cardioverter-defibrillator implantation and embolization of the left anterior descending artery followed by right ventricular pacing (240 ppm) for 3 weeks to produce heart failure. In stage 2, 28 surviving animals were assigned to the SCS (delivered at the T4/T5 spinal region for 2 hours 3 times a day), medicine (MED; carvedilol therapy at 12.5 mg PO BID), or control (CTRL; no therapy) group for the initial phase 1 study. In a subsequent phase 2 study, 32 stage 1 survivors were equally randomized to the SCS, MEDS (carvedilol plus ramipril 2.5 mg PO QD), SCS plus MEDS (concurrent therapy), or CTRL group. Animals were monitored for 5 weeks (phase 1) or 10 weeks (phase 2). In stage 3, all phase 1 animals underwent circumflex artery balloon occlusion for 1 hour. In the SCS group, left ventricular ejection fraction was 65+/-5% at baseline, 17+/-3% at the end of stage 1, and 47+/-7% at the end of stage 2. In the MED group, left ventricular ejection fraction was 61+/-4% at baseline, 18+/-3% at the end of stage 1, and 34+/-4% at the end of stage 2. In the CTRL group, left ventricular ejection fraction was 64+/-5% at baseline, 19+/-5% at the end of stage 1, and 28+/-3% at the end of stage 2. Left ventricular ejection fraction was significantly improved in the SCS compared with the MED and CTRL groups (P<0.001 for both). The mean number of spontaneous nonsustained ventricular tachyarrhythmias during stage 2 and the occurrence of ischemic ventricular tachyarrhythmias during stage 3 also were significantly decreased in the SCS (27+/-17 and 27%, respectively; P<0.03) and MED (58+/-42 and 33%; P<0.05) versus CTRL (88+/-52 and 76%) group. After 10 weeks in the phase 2 studies, the greatest recovery in ejection fraction was noted in the SCS (52+/-5%) and SCS+MEDS (46+/-4%) groups compared with the MEDS (38+/-2%) and CTRL (31+/-4%) groups., Conclusions: SCS significantly improved cardiac contractile function and decreased ventricular arrhythmias in canine heart failure.