Objective To investigate the value of thromboelastography ( TEG) in evaluating the prognosis of polytrauma patients. Methods A retrospective case control study was conducted to analyze the clinical data of 155 polytrauma patients admitted to 94th Hospital of People's Liberation Army from September 2015 to December 2017. There were 118 males and 37 females, aged 18-88 years [(49. 0 ± 1. 3)years]. Injury severity score (ISS) was (26. 4 ± 11. 0)points. According to the prognosis of patients 90 days after injury, the patients were divided into survival group ( 143 patients ) and death group (12 patients). The ISS on admission, acute physiology and chronic health evaluation Ⅱ (APACHE Ⅱ), Glasgow coma score ( GCS ) , systolic blood pressure ( SBP ) , respiratory rate ( RR ) and outcome of 90 days after injury were collected. Red blood cells( RBC) , hemoglobin( Hb) , plasma prothrombin time ( PT) , activated partial thromboplastin time ( APTT) , fibrinogen ( FIB) , international normalized ratio ( INR) , thrombin time ( TT) , platelet count ( PLT) and TEG-related indicators within 2 hours after ICU admission were collected. TEG-related indicators including TCG clotting reaction time ( R ) , clot formation rate (K), clot formation kinetics (α angle), maximum amplitude (MA), coagulation index (CI), blood clotting strength (G), elasticity constant (E) and clot lysis time (CLT) were collected within 2 hours after admission to ICU. The changes of TEG-related indicators were compared between the two groups, and logistic regression analysis and receiver operating characteristic ( ROC) curve analysis were performed. Results Compared with the survival group, ISS and APACHE II scores in the death group were significantly higher, while GCS, SBP, RR, RBC and Hb were significantly lower (P<0. 05). Compared with the survival group, death group showed significantly prolonged PT [(21. 1 ± 9. 1) s vs. (16. 5 ± 4. 3)s] as well as reduced FIB [(1. 7 ± 1. 5)g/L vs. (2. 7 ± 1. 7)g/L] (P<0. 05). Compared with the survival group, the K value in the death group was significantly prolonged [5. 2(1. 8,8. 0)min vs. 2. 9(2. 1,4. 2)min] (P<0. 05). Compared with the survival group, death group showed significant decrease in G value (3762. 4 ± 3346. 7 vs. 6366. 5 ± 3886. 1), E value (75. 3 ± 67. 0 vs. 127. 3 ± 77.7),αangle[(38.4±19.7)°vs. (49.4±16.6)°],MA[(37.4±17.5)mmvs.(52.0±13.3)mm], CI [-6. 8(-9. 5, 1. 5) vs. -2. 9 ( -5. 9, -0. 7)] and CLT [39. 2 (32. 5, 46. 9) min vs. 46. 4 (32. 7, 60.8) min] (P<0. 05). There were no significant differences in APTT, TT, INR, PLT and R between the two groups (P>0. 05). Logistic regression analysis showed that MA was significantly associated with the prognosis of polytrauma patients (OR=1. 15, 95％ CI 1. 04-1. 28, P<0. 05). The area under the ROC curve of MA was 0. 756 (P <0. 05). When the MA threshold was 43. 1 mm, the sensitivity, specificity, positive predictive value and negative predictive value of predicted survival were 77. 5％, 76. 9％, 76. 1％ and 97. 3％, respectively. Conclusion TEG index MA can determine the prognosis of polytrauma patients, and the mortality rate is significantly increased at MA<43. 1 mm.