Abstract: In order to assess the variation of maize growth due to climate change, the maize yield model is upgraded and coupled with a regional climate model named PRECIS. The maize growth period and yield in Northeast China are simulated both under baseline (1961-1990) and B2 climate change scenario (2011-2050). The variations over the next 40 years are predicted by considering and not considering CO₂ fertilizer efficiency (direct influence) separately. A direct influence module of CO₂ is added into the maize growth model to make concentration of CO₂ as an input variable. The upgraded model can simulate the yield and the increase of C4 crop, especially maize, with different concentration of CO₂. And results fit the field experiments well. Furthermore, this model could distinguish fertilizer efficiency of photosynthesis and transpiration. Under B2 scenario, the temperature rises continuously and is higher than the baseline (1961-1990). The precipitation is less on the whole, and the radiation is more than the baseline. What calls for special attention is that the precipitation is more in the 2020s, which is favorable to maize growth. Without considering CO₂ fertilizer efficiency, the production almost decreases, and the range of reduction closely relates to maturity. The reduction is biggest in parts of Songnen Plain, more than 20%. But in the 2020s, the production in most areas increases less than 20%. The variation is caused by weather condition, and the increasement of temperature and decreasement of precipitation should be the primary cause. As time goes on, the reduction is bigger and bigger. In the 2020s, the precipitation always is greater and beneficial to the maize growth. The CO₂ fertilizer efficiency is important, and its compensation effects on the maize yield is significant. The distribution of the yield variation is similar, but the range is less. As the concentration of CO₂ goes higher, the CO₂ fertilizer efficiency is more and more significant. So, the CO₂ fertilizer efficiency and weather condition must be considered. In the next 40 years, the variation of the maize growth period distributes relatively stable, and closely relates to the maturity. The growth period of mid-and early-maturation shorten obviously, but that of late-maturation elongate persistently. Changes of other three maturities are not obvious.The variation of yield and growth period has theoretical significance comparing with the average value of 1961-1990. But these results are based on the consideration of climate change barely, without social feedback and adaption to climate change, so the model needs improving in the future.