This paper designs a predictive speed controller of interior permanent-magnet synchronous motor (IPMSM) based on finite control set (FCS)-model predictive control (MPC). The proposed predictive speed controller has a cascade-free structure that comprises a single predictive control function for the speed and dq-axis stator currents. The future rotor speed and stator currents are predicted at each sampling interval. Then, the proposed predictive control function is evaluated to determine the optimal switching states for the IPMSM drive. In high-performance speed-controlled drive, the unknown load torque disturbance has to be suppressed. Therefore, the load torque disturbance estimator using the two-stage extended Kalman filter (TSEKF) is proposed to enhance the dynamic performance of the IPMSM drive. The proposed TSEKF simultaneously estimates the system states, which greatly improve the model prediction process. The proposed predictive speed control is simple and compact, which demonstrates a faster transient response and a robust feature to mechanical system parameter variations when compared with the conventional cascaded speed control under a wide speed range and load torque variations. The experimental results on a prototype IPMSM drive built on a Texas Instruments TMS320F28335 floating point digital signal processing (DSP) board are presented considering the constant torque and flux-weakening regions to confirm the feasibility of the proposed FCS-MPC scheme.