Objective: To conduct a comprehensive review of risk prediction models for oral mucosal pressure injury in ICU patients with oral tracheal intubation, aiming to provide a reference for the development and clinical application selection of risk prediction models in the future. Methods: Using the scope review methodology framework, this study searched PubMed, Web of Science, Embase, Cochrane Library, China Biomedical Literature Database, CNKI and Wanfang from database construction until December 31. Two researchers independently screened and extracted data for aggregated analysis. Results: Seven studies were finally included. The subjects were mainly ICU patients with oral tracheal intubation. Logistic regression model was used as the model construction method, and nomogram was used as the model presentation. The three most frequently occurring predictors were tracheal intubation indwelling time, APACHEⅡ score, and dental pad usage. Conclusion: The included models had good overall predictive power. However, to establish models with a low risk of bias, good predictive performance, and strong operability, a high overall risk of bias needed to be addressed to provide some guidance for clinical practice.