This paper describes a planetary rover tracking control method by estimating the slip rate based on the echo state network (ESN). A rover is required to move on uneven terrain that varies depending on the location. Soft soils should be particularly avoided during planetary rover exploration. However, it is not easy to detect them with external sensors such as cameras and light detection and ranging (LiDAR). The rover would be stuck on soft soils because of the high slip rate if the thrust of the rover is determined without considering the slip rate. Therefore, estimating the slip rate at the current position is necessary to generate the appropriate thrust. We propose the method to estimate the slip rate that varies depending on the location with a few sensors. The proposed method uses the ESN to compensate for the effect of thrust loss caused by the slip rate. The ESN is one of the recurrent neural networks and can estimate some values quickly using time series data of inertial measurement unit (IMU). The ESN is used to estimate the slip rate of the current rover position in real time. Results of the numerical simulation show the effectiveness of the proposed tracking control method for the exploration rover.