Abstract: Deadbeat predictive current control （DPCC） is widely applied to permanent magnet synchronous motors （PMSM） due to its rapid dynamic response and excellent tracking performance. However, for the speed-current dual closed-loop control systems, the traditional DPCC uses proportional-integral （PI） control in the speed outer loop, which has limited disturbance suppression capability. Moreover, DPCC relies on an accurate PMSM mathematical model, and parameter mismatches between the actual motor and the original model under certain conditions can degrade control accuracy. A dual-observer-based disturbance suppression DPCC method for PMSM was proposed to solve the above problems. First, a sliding mode control （SMC） combined with a sliding mode load torque observer （SMLTO） method was employed in the speed outer loop. The SMLTO can track load torque variations in real time and convert them into current values to compensate for the reference current calculated by SMC. Then, A modified error feedback function-based extended state observer （MESO） was employed in the current inner loop, where the proposed error feedback function features lower computational complexity. The MESO can estimate unknown disturbances caused by parameter mismatches and compensate for the reference voltage in the DPCC method. Finally, simulation and experimental results were conducted to verify the effectiveness of the proposed method.