Objective: The present study was designed to assess in vivo a new, optimized, virtually wall-less, dual-lumen, bi-caval cannula for veno-venous ECMO in comparison to a commercially available cannula. Methods: Veno-venous extracorporeal membrane oxygenation (ECMO) was carried out in a bovine study (n=5, bodyweight 75±5kg). Following systemic heparinization, ECMO was established in a trans-jugular fashion through a calibrated 23F orifice, using a new, optimized, virtually wall-less, dual-lumen, bi-caval 24F cannula (Smartcanula LLC, Lausanne, Switzerland) versus a commercially available 23F bi-caval, dual-lumen control cannula (Avalon Elite^®, Maquet, Rastatt, Germany) in a veno-venous ECMO setup. Veno-venous ECMO was initiated at 500 revolutions per minute (RPM) and increased by incremental steps of 500 RPM up to 2500 RPM. Catheter outlet pressure, catheter inlet pressure, oxygen saturation and pump flow were recorded at each stage. Results: Mean flow accounted for 0.37±0.04 L/min for wall-less versus 0.29± 0.07 L/min for control at 500 RPM, 0.97±0.12 versus 0.67±0.06 at 1000 RPM, 1.60±0.14 versus 1.16±0.08 at 1500 RPM, 2.31±0.13 versus 1.52±0.13 for 2000 RPM and 3.02±0.5 versus 2.11±0.18 (p<0.004). The mean venous suction required was 19±8 mmHg for wall-less versus 20±3 mmHg for control at 500 RPM, 7±3 versus 9±4 for 1000 RPM, -11±10 versus -12±8 at 1500 RPM, -39±15 versus -49±10 for 2000 RPM and -60±28 versus -94±7 for 2500 RPM. The mean venous injection pressure accounted for 29±7 mmHg for wall-less versus 27±5 mmHg for control at 500 RPM, 50±6 versus 61±7 at 1000 RPM, 89±10 versus 99±17 for 1500 RPM, 142±14 versus 161±9 at 2000 RPM and 211±41 versus 252 ±3 for 2500 RPM. Conclusion: Compared to the commercially available control cannula, the new, optimized, virtually wall-less, dual-lumen, bi-caval 24F cannula allows for significantly higher blood flows, requires less suction and results in lower injection pressures in vivo.