Objective:To construct a prediction model for the emergency volume of the psychiatric hospital, and analyze the changes of psychiatric emergency visits, so as to provide references for optimizing the allocation of emergency service resources.Methods:This study extracted data of the visit time of emergency patients, etc from the information system of a certain psychiatric hospital from 2018 to 2023. The monthly emergency visits (emergency volume) from 2018 to 2022 were used to construct the autoregressive integrated moving average model (ARIMA), and the monthly emergency volume from 2023 was used to validate the predictive performance of the model.Results:After model construction and screening, seasonal ARIMA (0, 1, 0) (1, 1, 1) 12 was determined as the optimal model. The predicted values of the model were in good agreement with the actual values, with an average relative error fluctuation of 1.6% to 26.8% and an average absolute error fluctuation of 9 to 159 person-time. Conclusions:The seasonal ARIMA model could accurately predict the emergency volume of a certain psychiatric hospital and provide references for human resource allocation and emergency response. However, this prediction model was suitable for short-term forecasting. If long-term forecasting was needed, continuous data fitting was necessary to ensure the effectiveness of the prediction.