This paper analyzed the relationship between the rice yields of Vietnam and Thailand, and global monthly average temperature of the sea surface and the height of 500 hPa. To separate the trends yield and meteorological yield, eleven different statistical methods were adopted. According to relevant widespread investigations, the regression models of different meteorological yield and the sea surface temperature and the circulation were built up by using step regression method based on the coefficient. The result of the tendency yield and meteorological yield estimated through growth curve method of Vietnam and Thailand are preferable by statistical testing. Based on the physical significance of the regression model, the yield in 2002 was predicted and the error was only 2%-5%. The results may be applied in practice.