To evaluate the safety and efficacy of argatroban applied as alternative anticoagulant in critical illness patients underwent extracorporeal membrane oxygenation (ECMO) with contraindications of unfractionated heparin (UFH), and to further explore the effective dose of argatroban. From July 1, 2013 to February 28, 2022, there were 14 patients who admitted in the respiratory intensive care unit (RICU) of Beijing Chao-Yang Hospital received ECMO and used argatroban for anticoagulation (argatroban group). Two of them received argatroban as the initial anticoagulant. The remaining 12 patients used UFH at first, and then switched to argatroban. UFH group included 28 patients who received UFH for anticoagulation after matching the demographic characteristics. Primary endpoint was the prevalence of ECMO-related thrombotic events. Secondary endpoints included the type of thrombotic events, prevalence of ECMO-related major bleeding events, bleeding sites, ICU mortality, mortality during ECMO, liver and kidney function, thrombelastogram, blood transfusion, dosage of argatroban, the dynamic changes of coagulation variables 4 days before and 7 days after argatroban treatment. In argatroban group, there were 8 patients received veno-venous ECMO (VV-ECMO), 2 patients with veno-arterial ECMO (VA-ECMO), and 4 patients with veno-arterio-venous ECMO (VAV-ECMO). In UFH group, VV-ECMO was applied in 23 patients, VA-ECMO and VAV ECMO was established in 3 patients and 2 patients, respectively. In endpoint events, the incidence of ECMO related thrombotic events in argatroban group was slightly higher than that in UFH group (28.6% vs. 21.4%). The ECMO running time in argatroban group was slightly longer than that in UFH group [days: 16 (7, 21) vs. 13 (8, 17)]. The incidence of ECMO-related bleeding events (28.6% vs. 32.1%) and mortality during ECMO (35.7% vs. 46.4%) in argatroban group were slightly lower than those in UFH group. However, the differences were not statistically significant (all P < 0.05). The platelet transfusion in argatroban group was significantly higher than that in UFH group [U: 7.7 (0, 10.0) vs. 0.8 (0, 1.0)]. The coagulation reaction time (R value) in thrombelastography in argatroban group was significantly longer than that in UFH group [minutes: 9.3 (7.2, 10.8) vs. 8.8 (6.3, 9.7)]. The maximum width value [MA value, mm: 48.4 (40.7, 57.9) vs. 52.6 (45.4, 61.5)] and blood clot generation rate [α-Angle (deg): 54.1 (45.4, 62.0) vs. 57.9 (50.2, 69.0)] in the argatroban group were significantly lower than those in the UFH group (all P < 0.05). The activated partial thromboplastin time (APTT) was prolonged after changing from UFH to argatroban in the argatroban group [seconds: 63.5 (58.4, 70.6) vs. 56.7 (53.1, 60.9)]. The PLT level showed a decreasing trend during UFH anticoagulation therapy, and gradually increased after changing to argatroban. D-dimer level was 19.1 (7.0, 28.7) mg/L after switching to argatroban, and then no longer showed an increasing trend. The level of fibrinogen (FIB) showed a decreasing trend during the anticoagulant therapy of UFH (the lowest was 23.6 g/L), and fluctuated between 16.8 and 26.2 g/L after changing to argatroban. The median initial dose of argatroban was 0.049 (0.029, 0.103) μg×kg-1×min-1, which the highest dose was in VV-ECMO patients of [0.092 (0.049, 0.165) μg×kg-1×min-1]. The initial dose of VAV-ECMO was the lowest [0.026 (0.013, 0.041) μg×kg-1×min-1], but without significant difference (P > 0.05). The maintenance dose of argatroban was 0.033 (0.014, 0.090) μg×kg-1×min-1, VV-ECMO patients was significantly higher than those in VA-ECMO and VAV-ECMO patients [μg×kg-1×min-1: 0.102 (0.059, 0.127) vs. 0.036 (0.026, 0.060), 0.013 (0.004, 0.022), both P < 0.05]. Argatroban appears to be a feasible, effective and safety alternative anticoagulant for patients with contraindications to UFH who undergoing ECMO support.