Abstract: Temperature determines the maturity of crops by affecting the formation of dry matter, and ultimately affects crop yield. Therefore, studying the relationship between temperature and crop growth is very important to improve agricultural production efficiency. In this study, an universal Logistic model for describing the growth process of rice was established with growing degree days as independent variable, the leaf area index and dry matter accumulation of rice as dependent variable, respectively. A large number of data of rice growth index were collected. At the same time, the relationships between the maximum leaf area index, the maximum dry matter accumulation, harvest index (the ratio of crop economic yield to total biomass) and precipitation throughout the growth period of rice were analyzed. The results showed that with the increase of growing degree days, the change of rice leaf area index revealed obvious characteristics： increase firstly then decrease, and the change trend of leaf area index in different regions was basically similar. When the growing degree days was increased to about 1 000 ℃, the leaf area index reached the maximum and this stage corresponded to the booting stage. When the growing degree days was increased to about 1 000 ℃, the increase rate of dry matter accumulation was the largest. At this time, the leaf area almost stoped growing, and the rice entered the reproductive growth stage. There was an obvious linear relationship between the maximum leaf area index and the maximum dry matter accumulation of rice. With the increase of the maximum leaf area index, the maximum dry matter accumulation increased gradually. There was a quadratic polynomial relationship between the maximum leaf area index and harvest index of rice. When the maximum leaf area index was about 8, the harvest index was the largest. The maximum leaf area index of rice increased first and then decreased with the increase of precipitation in the whole growing period. When the precipitation of the whole growing period was about 670.5 mm, the maximum leaf area index of rice increased to about 7.9. The corresponding harvest index was also almost the maximum value. If the precipitation was more than 670.5 mm, the growth of rice leaves would be inhibited, and the photosynthesis of rice leaves would be weakened, so that the function of leaves could not be brought into full play. At the same time, too much precipitation was not conducive to the growth of rice roots, resulting in yield reduction. In this study, the rice growth regulation and yield prediction model were constructed. The results showed that the growing degree days could be used to analyze the growth process of rice accurately, and it could improve the precision of rice yield prediction and efficiency agricultural production. In this study, we considered the relationsips between the meteorological factors and the crop growth, and established the universal rice growth model by using growing degree days, precipitation and the physiological index of the rice, and estalished the prediction model of the rice yield by using the harvest index. It would be a guidance of constructing growth models of other crops.