BACKGROUND: Although venoarterial extracorporeal membrane oxygenation (VA-ECMO) is beneficial for the treatment of profound cardiogenic shock, peripheral VA-ECMO cannulation can increase left ventricular afterload, thus compromising myocardial recovery. We investigated whether early routine left ventricular unloading can reduce 30-day mortality compared with the conventional approach in patients with cardiogenic shock undergoing VA-ECMO. METHODS: This randomized clinical trial involved 116 patients with cardiogenic shock undergoing VA-ECMO from March 2021 to September 2022 at Chonnam National University Hospital, Gwangju, South Korea. The patients were randomly assigned to undergo either early routine left ventricular unloading with transseptal left atrial cannulation within 12 hours after randomization (n=58) or the conventional approach, which permitted rescue transseptal left atrial cannulation in case of an increased left ventricular afterload (n=58). The primary outcome was all-cause mortality within 30 days. RESULTS: All 116 randomized patients (mean age, 67.6±13.5 years; 34 [29.3%] women) completed the trial. At 30 days, all-cause death had occurred in 27 (46.6%) patients in the early group and 26 (44.8%) patients in the conventional group (hazard ratio, 1.02 [95% CI, 0.59–1.74]; P =0.942). Crossover to rescue transseptal left atrial cannulation occurred in 29 patients (50%) in the conventional group according to a clear indication. Time to rescue transseptal cannulation in the conventional group was a median of 21.8 (interquartile range, 12.4–52.2) hours after randomization. There were no significant differences in other secondary outcomes between the 2 groups except for a shorter time to disappearance of pulmonary congestion in the early group (median, 3 [interquartile range, 2–6] versus 5 [interquartile range, 3–7] days; P =0.027). CONCLUSIONS: Among patients with cardiogenic shock undergoing VA-ECMO, early routine left ventricular unloading with transseptal left atrial cannulation did not reduce 30-day mortality compared with the conventional strategy, which permitted rescue transseptal left atrial cannulation. These findings should be cautiously interpreted until the results of multicenter trials using other unloading modalities become available. REGISTRATION: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT04775472.