Abstract: Abstract: Ethanol production from energy crops which are renewable resources has gotten more and more attentions because of the energy crisis and environmental pollution. Sweet sorghum is considered as the most promising energy crop for the production of ethanol. Sweet sorghum stem is usually used to ferment ethanol, one of which is the liquid-state fermentation with the juice of sorghum stem. But a lot of bagasse of sorghum stem is discarded as wastes. The bagasse can be used as the supplemental materials of the fermentation of the juice. So it is absolutely necessary to study on the optimization of ethanol production with the bagasse and juice of the sorghum stem by simultaneous saccharification and fermentation (SSF). The response surface is an effective method to optimize the operating parameters of the SSF for the maximum ethanol yield. In this study, the Plackett-burman design was adopted to select the significant factors from 8 variables which influenced the ethanol yield and its concentration. The results indicated the ethanol yield and concentration were mainly influenced by the fermentation temperature, the amount of cellulase and the ratio of sorghum stem bagasse to its juice (P<0.05). And the other 5 variables which were not significant were remained to be the center level: the pH value was 5.5, the inoculation ratio was 0.2%, the amount of (NH4)2SO4 was 5 g/L, the fermentation time was 60 h, and the amount of uracil was 1 g/L. Based on the results of the selection, the steepest ascent experiment was conducted to determine the center point and the step size. The center point of the fermentation temperature was 37℃ and its step size was 3℃. The center point of the amount of cellulase was 20 FBU/30 CBU and its step size was 5 FBU/CBU. The center point of the ratio of bagasse to its juice was 7% and its step size was 1.5%. Then we adopted the Box-Behnken design and response surface analysis method to optimize the levels of these 3 significant factors, and the regression and the optimal levels of these significant factors were as follows: the regression model was very significant (P<0.0001, R2=0.9987, Adj. R2=0.9971 and CV=0.0099), which indicated that the model was reliable, and therefore, the regression model could be used for the theoretical prediction of the SSF for juice and bagasse of sweet sorghum stem; the fermentation temperature was 36.58℃, the amount of cellulase was 23.5 FBU/35.25 CBU, the ratio of bagasse to juice of sorghum stem was 8.2%, the maximum theoretical yield of ethanol predicted was 89.2%, and the ethanol concentration was 31.829 g/L. The interactive effects were analyzed by response surface analysis. The interactive effect of the fermentation temperature and the amount of cellulase was very significant (P<0.0001); the interactive effect of the amount of cellulase and the ratio of bagasse to juice was very significant (P<0.001); the interactive effect of the fermentation temperature and the ratio of bagasse to juice was not significant (P=0.2946). Under the optimal conditions, the model was proved to be valid by the verification test with 3 repeated tests, and the ethanol yield was 88.98% which was very close to the maximum yield (89.2%). The ethanol concentration was 31.78 g/L, which was very close to theoretical prediction value (31.829 g/L). The results of this optimal technology will provide a reference for the ethanol production technology by simultaneous saccharification and fermentation method for sweet sorghum stem.