By using the historical data of rice planthopper in Guilin of Guangxi and related meteorological data, the factors puffing technology was employed to assemble the meteorological data in pentad scale. Through the correlation analysis of the meteorological data and the occurrence of rice planthopper, the key factors and their assembling periods most significantly related with the occurrence of rice planthopper were screened. Two meteorological forecast models for the occurrence grade of rice planthopper in pentad scale were built, of which, one applied multiple linear regression model was built by using SPSS software, and the another applied BP artificial neural network model was built by using MATLAB software. Based on the factors puffing technology, the prediction result as “correct” and “basically correct” of the two models were above 88%. The prediction results as “correct” and “basically correct” of the artificial neural network model was increased by 4% and the prediction result as carrect was increased by 14%, according to the historical samples and the independent samples, respectively. Therefore, a short-term forecasting model built with factors puffing technology and BP artificial neural network model could not only realize the dynamic updating and forecasting of the occurrence grade of rice planthopper, but also evidently improve the prediction accuracy and stability.