【Objective】The present study was conducted to establish a dynamic forecasting model for occurrence of rice leaf roller， and to provide scientific basis for its control. 【Method】Three meteorological factors which had a close degree and better correlation with the 3rd generation of rice leaf roller were evaluated using the principle of Fuzzy closeness in combination with the actual conditions of the 3rd generation of rice leaf roller in Huazhou City， the prediction guideline for Fuzzy closeness was set. 【Result】The quantitative prediction model equation for 3rd generation of rice leaf roller occurrence in Huazhou City was established as y=2.5587+44.1039 R（x0， xi）， and the prediction values were calculated. The rate of matching results with historical data reached 92.3%， and the predicted results was coincided with the actual data taken in the years 2009 and 2010 which revealed that the composition and structure of prediction model factors of the 3rd generation of leaf roller population met with the prediction requirements. 【Conclusion】The established model was convenient and involved simple calculations and lesser labour， and gave results with high precision rate. It is suitable for popularization and application at primary report stations and may be used in the production units.