BACKGROUND: During venovenous extracorporeal membrane oxygenation (vvECMO), direct thrombin inhibitors are considered by some potentially advantageous over unfractionated heparin (UFH). We tested the hypothesis that Argatroban is non-inferior to UFH regarding thrombosis and bleeding during vvECMO. METHODS: We conducted a propensity-score matched observational non-inferiority study of consecutive patients without heparin-induced-thrombocytopenia (HIT) on vvECMO, treated between January 2006 and March 2019 in the medical intensive care unit at the University Hospital Regensburg. Anticoagulation was realized with UFH until August 2017 and with Argatroban from September 2017 onwards. Target activated partial thromboplastin time was 50 ± 5seconds in both groups. Primary composite endpoint was major thrombosis and/or major bleeding. Major bleeding was defined as a drop in hemoglobin of ≥ 2 g/dl/day or in transfusion of ≥ 2 packed red cells/24 h, or retroperitoneal, cerebral, or pulmonary bleeding. Major thrombosis was defined as obstruction of > 50% of the vessel lumen diameter by means of duplex sonography. We also assessed technical complications such as oxygenator defects or pump head thrombosis, the time-course of platelets, and the cost of anticoagulation (including HIT-testing). RESULTS: Out of 465 patients receiving UFH, 78 were matched to 39 patients receiving Argatroban. The primary endpoint occurred in 79% of patients in the Argatroban group and in 83% in the UFH group (non-inferiority for Argatroban, p = 0.026). The occurrence of technical complications was equally distributed (Argatroban 49% vs. UFH 42%, p = 0.511). The number of platelets was similar in both groups before ECMO therapy but lower in the UFH group after end of ECMO support (median [IQR]: 141 [104;198]/nl vs. 107 [54;171]/nl, p = 0.010). Anticoagulation costs per day of ECMO were higher in the Argatroban group (€26 [13.8;53.0] vs. €0.9 [0.5;1.5], p < 0.001) but not after accounting for blood products and HIT-testing (€63 [42;171) vs. €40 [17;158], p = 0.074). CONCLUSION: In patients without HIT on vvECMO, Argatroban was non-inferior to UFH regarding bleeding and thrombosis. The occurrence of technical complications was similarly distributed. Argatroban may have less impact on platelet decrease during ECMO, but this finding needs further evaluation. Direct drug costs were higher for Argatroban but comparable to UFH after accounting for HIT-testing and transfusions.