Abstract: Mycosphaerella larici-leptolepis lto et al increasingly occurs across global forest in recent years with the climate change. To effectively prevent large outbreak of the disease, two models were constructed by the authors to monitor and forecast the occurrence of the disease via meteorology data analysis. The results showed that Mycosphaerella larici-leptolepis lto et al could easily survive in the climate condition of low temperature with high humidity, but was venerable to high temperature with low humidity. Results from the two models of stepwise regression and neural network showed that the average error of stepwise regression model and neural network model was estimated to 18.8% and 8.6% , respectively. In addition, average errors generated from stepwise regression model that tested by meteorology data of 2012 and 2013 were estimated to 7.2% and 11.4% while average errors generated from neural network model were estimated to 5.3% and 2.4% , respectively. Errors obtained from neural network model were 1.9% and 9.0% lowered than that from stepwise model, respectively. The results suggested that the neural network model exhibited better performance in monitoring the disease compared with the stepwise regression model, and could be employed in scientific forecasting of Mycosphaerella larici-leptolepis lto et al in forest of northeast China.