In this paper, a model predictive control algorithm based on dead-zone time optimal compensation is proposed for permanent magnet synchronous motor. First, the dead-time voltage vectors corresponding to each candidate voltage vectors are analyzed. And a new finite control set is obtained under the minimum dead-zone time. Based on the new control set, the predictive controller is designed to obtain the optimal voltage vector and its dead-time voltage vector. Then the beneficial dead-time voltage vector is judged, and the dead-zone time is optimized. The optimal dead-zone time and the optimal voltage vector are used to achieve speed stability and good performance. Finally, a simulation example is established to verify the feasibility and effectiveness of the proposed algorithm.