Rehabilitation robots are advanced systems designed to assist injured individuals by providing targeted support and facilitating therapeutic exercises, thereby helping to improve physical functions following injuries or medical interventions. This study focuses on the modeling and controller design of a rehabilitation robot. In the modeling section, the robot is considered as a planar three-degree-of-freedom serial robot to enable the actuation of the hip, knee, and ankle joints. The dynamic model of the robot is derived using the Euler–Lagrange method. Furthermore, with the increasing emphasis on robot-assisted rehabilitation, the need for adaptive control systems to tailor treatment to the individual needs of patients has become critically important. Therefore, due to the variability and uncertainty of the robot parameters, an adaptive controller is employed. In addition, considering the presence of noise in sensor data and disturbances arising from the interaction between the robot and the patient, a high-gain observer is used to estimate the system state variables, and a disturbance observer is utilized to estimate and compensate for disturbance effects. The obtained results demonstrate that the designed control system is capable of tracking the reference input with an error of less than 0.05 radians.