With Japonica cultivar Wuxiangjing 14 as test material, a free-air CO2 enrichment (FACE) experiment was conducted at Anzhen and Wuxi of Jiangsu Province in 2001 to approximately 2003. The target CO2 concentration of FACE plots was 570 micromol x mol(-1), 200 micromol x mol(-1) higher than that of ambient air. Three levels of N were supplied as LN (150 kg x hm(-2)), MN (250 kg x hm(-2)) and HN (350 kg x hm(-2)). The effects of FACE treatment on the dynamics of rice tiller was studied, and the simulation model was established as Tt = ((A1/1 + e(a1-b1t)) - (A2/1 + e(a2-b2t)) + C) x ((B1/(1 + e(a3-b3t) - (B2/1 + e(a4-b4t))+D) Where Tt was the numbers of rice tiller in t days after transplanting; A1 and A2 were the maximal tillers of production and death under ambient air, respectively; B1 and B2 were the maximal potential tillers of production and death under FACE, respectively; C was the coefficient of CO2 concentration; D was the initial tillers after transplanting; and a1, a2, a3, a4, b1, b2, b3 and b4 were the control coefficients of the model. The dynamics of tiller numbers with the days after transplanting was described, and the model fitted well under ambient air and FACE conditions. Through testing with different year experimental data, the maximum and minimum RMSE was forecasted as 44.27 and 13.96 ind x m(-2), respectively, suggesting that the model was accurate and applicable.