Viscoelastic tests allow a reduction in blood product transfusion. Three modern devices are currently available (rotational thromboelastometry [ROTEM] sigma, thromboelastography [TEG] 6S, and Quantra). No study has compared the performances of these 3 devices simultaneously. An observational, nonrandomized cohort study. A single-center of cardiac surgery in a university hospital. A total of 30 consecutive measurements from at least 10 adult patients presenting significant bleeding in the intensive care unit after cardiac surgery INTERVENTION: Viscoelastic tests using ROTEM sigma, TEG 6S, and Quantra were performed concomitantly with conventional coagulation measurements MEASUREMENTS AND MAIN RESULTS: The authors included 16 patients with 31 blood samples. After the exclusion of missing values, 27 samples were analyzed. Correlation with platelet count was as follows: ROTEM, r=0.84 [0.66-0.93], p < 0.0001; Quantra, r=0.83 [0.64-0.92], p < 0.0001; TEG 6S, r=0.64 [0.29-0.83], p=0.001. Correlation with fibrinogen (Clauss assay) was as follows: ROTEM, r=0.85 [0.68-0.93], p < 0.0001; Quantra, r=0.88 [0.74-0.95], p < 0.0001; TEG 6S, r=0.79 [0.55-0.91], p < 0.0001. No difference was observed for the detection of residual circulating heparin (anti-Xa activity >0.1), with 87% of correct identification for Quantra and 80% for both ROTEM and TEG 6S (p=0.3). Time to first results after the beginning of the test was shorter for Quantra than ROTEM and TEG 6S (136 [126-152] seconds v 205 [176-221] seconds, p=0.003 and v 450 [372-516] seconds, p < 0.0001 respectively). ROTEM sigma, TEG 6S, and Quantra performed similarly for exploring platelet count or residual circulating heparin. Thromboelastography 6S presented a weaker correlation with fibrinogen Clauss.