Abstract: Abstract: Global warming is a concern for policy makers, scientists and farmers around the world. It is widely accepted that climate change has had great impacts on agricultural production and farmers' income. Thus, there is a conundrum of how to balance the tradeoffs between food demand, increasing population, and scarce agricultural resources in the changing climate. Sensitivity and vulnerability of crop production to climate change are highly important for agriculture adaptation, but uncertainty still exists in previous research among diverse regions. Rice production may face huge risk because of the frequent extreme weather in China, therefore, a sensitivity analysis of rice production under the impacts of climatic change in different growth stages will contribute to optimizion of rice cultivation management for adapting to the changing conditions. In this study, we selected the recorded rice phonological observations at 10 agro-meteorological experiment stations, and the corresponding weather and statistical yields from 1980-2012 in Hunan Province, China. These data were used to analyze rice sensitivity to climate change using a panel model combined with multiple regression methods. The relationship of climate and yield trend was computed with the least square method, and the possible relationship between climatic factors and double-rice yield was analyzed by partial correlation analysis. The results showed that the double-rice region in Hunan Province experienced a warming trend over the last three decades, and the average temperature during the early rice and late rice seasons were 0.47 and 0.33 oC/(10a) higher. An increase in temperature of 0.76 oC/(10a) was observed during the vegetative stage of the early rice season, while the late rice season experienced a relatively slower increase in temperature. Precipitation and radiation during the growth stage of early rice tended to increase unnoticeable, but radiation in the vegetative stage and the reproductive stage of early rice changed with the trend -0.40 and 0.40 MJ/(m2·d·10a), while the trend was opposite for the late rice. Yield changes in the early rice were significantly correlated with precipitation and radiation during the ripening stage and the whole growth stage (P<0.01), but only the average temperature was significantly correlated to the yield for the late rice (P<0.05). Whatever the model adopted, the yield sensitivity of double-rice to climate change showed great differences among different growth stages, ranging from -280.11 to 118.20 kg/hm2. Temperature, precipitation and radiation in the vegetative stage were most sensitive to the yield for both the early rice and late rice. The rising temperature increased yield of the early rice, while we observed the opposite effect for the late rice. From 1980 to 2012, temperature increase, decrease of precipitation and radiation changed the early rice yield by -0.24% to 3.18%, and temperature increase during vegetative stage increased the early rice yield by 2.63%, and even accelerated the early rice yield by 157.10 kg/hm2. The late rice experienced a severe yield decrease by -4.98% with only temperature change in consideration. The historical climatic change impacted the yield of the early rice and late rice by 2.59% and -6.02%, respectively. These results suggest that the historical climate change during different rice growth stages has changed yields of double-rice in Hunan Province, especially if temperature rose during the growth stages. There is a strong need to assess historical climate change to project possible impacts in the future and identify adaptive technologies for double-rice production in this region.