To ascertain the key meteorological factors and their periods of influencing rice planthopper occurrence levels and provide early warning for local government and related departments in Guangdong Province, the relation between the rice planthopper occurrence areas and corresponding surface meteorological data from 1992 to 2009 was researched by correlation analysis method, and the long-term predicting models of rice planthopper occurrence levels were constructed by least-squares method and curve regression in June and September. The results indicated that June and September were the key periods of rice planthopper occurrence. key meteorological factors included the ratio of precipitation to air temperature, heavy rain days, days of average wind speed<5 m/s in June, and average temperature, days of average wind speed<5 m/s in September. The accuracy of rice planthopper occurrence level models was 83%, and prediction of rice planthopper occurrence level was correct in 2010. The prediction models had high prediction accuracy and could satisfy the needs of operational services.