INTRODUCTION: Adolescent suicide risk, particularly among individuals with depression, is a growing public health concern in China, driven by increasing social pressures and evolving family dynamics. However, limited research has focused on suicide prediction models tailored for hospitalized Chinese adolescents with depression. This study aims to develop a suicide risk prediction model for early identification of high-risk individuals using internal validation, providing insights for future clinical applications. METHODS: The study involved 229 adolescents aged 13-18 diagnosed with depression, admitted to a hospital in Shanxi, China. Feature selection was performed using Least Absolute Shrinkage and Selection Operator (Lasso) regression, and key predictors were incorporated into a multivariate logistic regression model. Model performance was assessed using the area under the receiver operating characteristic curve (AUC), Hosmer-Lemeshow test, calibration curves, decision curve analysis (DCA), and clinical impact curves (CIC). RESULTS: The model demonstrated AUC values of 0.839 (95% CI: 0.777, 0.899) for the training set and 0.723 (95% CI: 0.601, 0.845) for the testing set, indicating strong discrimination capability. Significant predictors included gender, social frequency, parental relationships, self-harm behavior, experiences of loss, and sleep duration. DCA and CIC supported the model's predictive potential. CONCLUSION: The model demonstrated strong predictive performance in internal validation, suggesting potential value for suicide risk assessment in hospitalized adolescents with depression. However, its generalizability remains to be confirmed. Further external validation in larger, multi-center cohorts is required to assess its robustness and clinical applicability.