The theory and technology of model predictive control(MPC) have been developed rapidly in recent years. With the successful applications of model predictive control for linear systems, predictive control for nonlinear models(NMPC)has received wide attention and achieved rich results. Based on the fundamental principle and characteristics of NMPC, in this paper, the current hot issues in the field and the results obtained were summarized, and it was point out that to study the NMPC for uncertainty systems and time-delay systems would be significant for further development of MPC theory and broadening MPC application fields.