Objective To investigate the clinical significance of thromboelastogram (TEG) and routine coagulation function monitoring in predicting bleeding in critically ill patients in the intensive care unit (ICU). Methods A retrospective analysis was performed for the clinical data of critically ill patients who were admitted to Department of Critical Care Medicine in our hospital from April 2021 to March 2023, and according to the presence or absence of bleeding, they were divided into bleeding group with 50 patients and non-bleeding group with 60 patients. All patients underwent the routine testing of four coagulation parameters and the TEG test, and the two groups were compared in terms of four coagulation parameters (prothrombin time [PT], activated partial prothrombin time [APTT], thrombin time [TT], and fibrinogen [FIB]) and TEG parameters (reaction time of coagulation [R], time to formation of blood clot formation time [K], α-angle, and maximum shear stress coefficient [MA]). A Spearman correlation analysis was used to investigate the correlation between the four coagulation parameters and TEG parameters; a logistic regression analysis was used to determine independent risk factors for bleeding; the receiver operating characteristic (ROC) curve was used to assess the predictive value of factors. Results Compared with the non-bleeding group, the bleeding group had a significantly higher PT value (P<0.05) and a significantly lower FIB value (P<0.05), and there were no significant differences in APTT and TT between the two groups (P>0.05). Compared with the non-bleeding group, the bleeding group had a significantly higher K value (P<0.05) and significantly lower α-angle and MA value (P<0.05), and there was no significant difference in R value between the two groups (P>0.05). R value was positively correlated with PT and APTT (P<0.05); K value was positively correlated with PT (P<0.05); α-angle was negatively correlated with PT (P<0.05); MA was negatively correlated with PT (P<0.05). The logistic regression analysis showed that PT, α-angle, and MA value were independent risk factors for predicting bleeding in critically ill patients in the ICU (P<0.05). The ROC curve analysis showed that the combination of these three parameters had the highest diagnostic value, with an area under the ROC curve of 0.924 (95% confidence interval: 0.857-0.966), a sensitivity of 81.82%, and a specificity of 92.73%. Conclusion The combined measurement of TEG and coagulation function has a good clinical application value in predicting the risk of bleeding and guiding platelet transfusion in critically ill patients in the ICU.