Abstract: BP neural network was used to simulate greenhouse micro-climate during rainy season in Jianghuai region of China. The effects of outside climate conditions and the degree of ventilation windows opening on greenhouse micro-climate was further analyzed. The results show that greenhouse microclimate during the rainy season can be simulated quite precisely with BP neural network, which can serves as a supplement of physical model for the greenhouse. During the rainy season, air humudity inside greenhouse is mainly affected by outside humidity, accounting up to 51.7% in weight among the five factors. Air flow rate inside greenhouse is mainly affected by outdoor wind speed and ventilation widow openings. Air temperature inside greenhouse is mainly dependent on outdoor temperature and solar radiation. The latter two factors account up to 46.2% and 27.9% in weight, respectively.