Abstract: Based on the data of 142 sample plots of Korean pine plantations in 7 areas of Changbai Mountain, the dominant climatic factors were screened according to Spearman correlation analysis and VIF value, and the optimal prediction model was determined by comparing the accuracy of linear model (LM) and generalized additive model (GAM). The production areas of Korean pine plantations in Changbai Mountain were divided by predicting the accumulation value at the base age. The results show that the average hottest monthly temperature (MWMT) and annual precipitation (MAP) are the main climatic factors affecting the accumulation and growth of Korean pine. In the fitting of GAM model, P spline (PS) is the optimal smoothing function, and its goodness of fit (AIC=618.37, -2LogL=604.82, R²=0.81) is significantly better than that of generalized linear model (LM). Ten-fold cross-validation proves that the prediction accuracy of GAM model meets the application requirements, and its stability is better than LM model. According to the prediction of GAM model, the study area can be divided into three grades: the most suitable, suitable and general growth area.