Abstract: The relationship between meteorological data and flowering data of peony in Mangshan Mountain from 1983 to 2010 was investigated. The results showed that flowering data was significantly correlated with the soil temperature and accumulated temperature. Through the selected meteorological factors， a prediction model was made by using SAS system. Furthermore， the relative error between the forecasted and the factual values ranged from -2.38% to 1.51%. The model could exactly forecast the florescence， which provided a reliable reference to predicting florescence of peony.

Key words: peony, florescence, climate, multiple linear regression, prediction model