The grey model theory is widely used in many field of investigation and of course includes inbound tourism demand. The original GM (1, 1) model couldn’t have accuracy prediction in some situation. Thus, an optimized GM (1, 1) model is proposed in this paper. Aiming at the deficiencies of the model, the improvements of GM (1, 1,) are include initial sequence, background value and parameter optimization. At last, the optimized GM (1, 1) model is used to predict inbound tourism demand of China and the results show that the proposed model is better than original GM (1, 1) model and time series model on prediction accuracy. © 2015 SERSC.