This study investigated the design of a drill bit stick–slip vibration and active disturbance rejection controller under the influence of an uncertain load at the bottom of a drill hole. There are two main problems in controlling the bit stick–slip vibration in a complex environment when stick–slip vibration occurs: the delay of the top drive output torque transmission and the uncertain load on the bit. Based on these two problems and the fact that the drilling system cannot be modelled accurately, in this study, we chose the model-independent active disturbance rejection control method to design the controller. The Smith predictor–based active disturbance rejection controller was designed by extending the state observer combined with the Smith predictor principle to achieve the delay compensation effect and introducing fuzzy online correction to cope with the robustness problem under uncertain load conditions, and a conventional Smith predictor–based proportional–integral–derivative controller was designed for comparison. It was verified by simulating oil drilling stick–slip vibration through simulation and semiphysical experiments. The experimental results show that the designed Smith predictor–based active disturbance rejection controller has better control performance, and the controller can effectively suppress the sudden change of torque and speed when a stick–slip vibration occurs, which can play a protective role for the safe and good operation of equipment.