Abstract: Rice phenological models are the bases of modern agro-meteorological service. Based on the conservation principle of crop physiological development stage, with the consideration of influences of climatic variables (i.e. temperature, sunshine hours and so on) on rice phenological stage, we built and verified the phenological forecasting model which were suitable for hsien rice and japonica rice planting region, using the observational data during rice phenology stage and ground meteorological data (2011-2014) from 12 agro-meteorological stations of Yunnan. The results showed that: both of prediction and observation value hsien and japonica phonological prediction models had good performances with 7.47 in averaged root-mean-square error (RMSE) and 7.99% in relative error (RE); the relative error (RE) was higher in hsien phonological prediction model (9.5%) than in japonica phonological prediction model (6.49%), implying the better performance of model prediction for japonica. Therefore, the biologically based and universally applied crop phonological process model is applicable for agrometeorological service.