AbstractThe effects of short‐term climate prediction are compared and analysed over China in summer 2003 using the coupled model with and without considering extra‐forcing factors (such as greenhouse gases, solar activity, and sulfate aerosols, etc). It is the fourth version coupled model, Global Ocean‐Atmosphere‐Land System model (GOALS 4.0), developed by State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics, Institute of Atmospheric Physics, Chinese Academy of Sciences. The simulations show that the prediction of precipitation distribution with variations of extra‐forcing is better than without extra‐forcing and closer to observations. In the simulation without extra‐forcing, precipitation is overestimated or underestimated in some areas of northern China, and the wet center is slightly shifted southward in southern China. But in the simulation with extra‐forcing, the pattern is more similar to the observation. For example, the strength of wet center in northern China is reduced and the center in southern China is boosted northward. The simulation successfully predicts the flood in the Huaihe River basin in summer 2003. Because of considering impact of extra‐forcing, the simulation can improve the case that the predicted precipitation is untrue in northern China, especially in Xinjiang. So, it should not be neglected to consider variations of extra‐forcing factors in short‐term climate prediction using the climate coupled model.