This article proposes a robust adaptive control scheme for robot manipulators by integrating sliding mode control, neural network approximation of uncertainties, and adaptive disturbance estimation. The proposed controller compensates for system uncertainties and unknown external disturbances commonly encountered in practical robotic applications. A radial basis function neural network approximates the unknown nonlinear dynamics, while an adaptive disturbance estimation is designed to approximate and reject external disturbances in real time. By incorporating these approximations into the control law, the proposed method significantly reduces the chattering phenomenon typically observed in conventional sliding mode control. Numerical simulations on a 3-degree-of-freedom robot manipulator confirm the controller’s ability to achieve accurate trajectory tracking while maintaining strong resistance to model uncertainties and external disturbances. Furthermore, the amplitude and frequency of chattering in the control signal are substantially reduced, confirming the improved performance of the control strategy. The results highlight the control approach’s reliability and practical applicability under uncertain conditions.