Abstract: The Institute of Atmospheric Physics Dynamical Climate Prediction System version 3.5 (IAP-DCPv3.5) was developed in this study based on the land–atmosphere coupled model component of the second-generation Chinese Academy of Sciences Earth System Model (CAS-ESM2.0). Specifically, we designed a land surface initialization scheme compatible with the Common Land Model and incorporated a stepwise regression-based bias correction method for precipitation prediction. Using 30-year ensemble hindcast experiment results from 1991 to 2020, we evaluated the system’s seasonal prediction skill for summer precipitation anomalies in China. Results demonstrated that IAP-DCPv3.5 exhibited reasonable predictive capability in eastern China, and the bias correction method significantly improved the system's prediction skills for summer precipitation. A comparison of prediction capabilities with different lead times suggested that the prediction skill for summer precipitation initialized in March was generally comparable to that initialization in May. An evaluation of prediction skills for summer monsoon circulation revealed that IAP-DCPv3.5 successfully predicted the interannual variation of the East Asian Summer Monsoon Index and reasonably reproduced the relationship between monsoon activity and precipitation anomalies in eastern China, indicating that its predictability for East Asian summer monsoon significantly affected its predictions of summer precipitation in China. Real-time prediction and validation of 2023 summer precipitation further demonstrated the system’s capability of predicting the observed spatial distribution of summer precipitation anomalies in China, despite discrepancies in the amplitude of predicted rainfall anomalies. This study highlights that IAP-DCPv3.5 can be directly applied for real-time summer rainfall prediction in China, and the validation of this system provides an important scientific evidence for further improving CAS-ESM2.0.