[Objective] To address the issue of large torque ripple in the direct torque control (DTC) system of brushless DC motor (BLDCM), this paper proposes a model predictive torque control (MPTC) method based on the open-winding brushless DC motor (OW-BLDCM). [Methods] First, a mathematical model of the OW-BLDCM was established, and the traditional voltage vector set was expanded by introducing a class of single-phase conducting voltage vectors to improve voltage adaptability. Next, based on the conventional DTC switching vector table, when torque needed to be increased, large vectors and single-phase vectors from the corresponding sector were used to increase torque. When torque needed to be reduced, in addition to the corresponding sector’s voltage vector, a model predictive control method was applied to roll and optimize the vector set, selecting the most effective voltage vector. Furthermore, simulation software was used to validate the proposed control method and the motor topology in terms of torque ripple suppression, observing the control effectiveness of torque and other parameters. Finally, physical experiments were conducted on an OW-BLDCM DTC platform. [Results] The results showed that the proposed MPTC method, compared to traditional DTC, achieved a 4.1% improvement in torque ripple suppression under rated operating conditions. The suppression of torque ripple was significantly enhanced at moderate and low speeds. Additionally, the proposed method in this study inherited the fast dynamic response characteristic, with a response time of only 1.706 ms, which was virtually identical to the 1.209 ms response time of DTC. [Conclusion] This study demonstrates that the proposed MPTC method for OW-BLDCM effectively reduces torque ripple and enhances control flexibility, providing a new approach for the efficient control of OW-BLDCM.