BackgroundIndividuals with schizophrenia are prone to higher rates of hospital readmission， presenting significant clinical challenges and imposing considerable social burdens within the mental health domain. In recent years， various risk prediction models have been developed to forecast readmission in patients with schizophrenia and support clinical decision-making， but their predictive performance and clinical applicability require comprehensive evaluation. ObjectiveTo systematically evaluate the risk prediction models for readmission in patients with schizophrenia， so as to provide insights for the development of high-performance and highly applicable readmission risk prediction models for patients with schizophrenia. MethodsOn July 5， 2025， a systematic literature search was conducted across multiple electronic databases， including PubMed， Embase， Cochrane Library， Web of Science， CINAHL， CNKI， China Biomedical Literature Database， Wanfang Database， and VIP Database， to identify risk prediction models for readmission in patients with schizophrenia. The search period was from the establishment of the databases to July 1， 2025. Two researchers independently performed literature screening， data extraction， risk of bias assessment， and applicability assessment. ResultsA total of 9 studies were included in this review， encompassing 18 risk prediction models for readmission in patients with schizophrenia. Among them， 4 models reported the area under the receiver operating characteristic （ROC） curve （AUC）， ranging from 0.734 to 0.820， 16 models provided AUC values of 0.642–0.879 for internal validation， and 1 model demonstrated an AUC of 0.841 for external validation. Key predictors included disease duration and the concomitant therapy of antipsychotic medications. The risk of bias was assessed as "high" in all included studies. ConclusionThe development of risk prediction models for readmission in patients with schizophrenia remains in an exploratory stage. Although the model exhibits favorable predictive performance， it is associated with a high risk of bias and insufficient performance evaluation.