An object is suspended by a magnetic field during the process of magnetic levitation. Magnetic Levitation Systems (Maglev) have considerable applications in numerous technical areas despite being unstable system. The sensitive, unreliable, and nonlinear nature of maglev systems makes it difficult to design reliable controllers. Adaptive controller with adaptation mechanism is developed as a solution to address the system uncertainties. In this thesis, First, the suspension control system model of the maglev train is built up, and the stability of the closed-loop system is analyzed. After that, A magnetic levitation system that is unstable is designed to be using Model Reference Adaptive control to track and control the train position. The performance of the controller is compared with PID and LQR. The simulation is done in Matlab2018a. The PID controller resulted performance characteristics overshoot (41%), settling time(24.87second), and rise time (1.04seconds) and the linear quadratic regulator controller having overshoot (19%), settling time(15.167second), and rise time (0.128seconds). MRAC resulted overshoot (0.39%), settling time(10.47second) and rise time (5.87second) It is generally good to compare this with the acceptable performance in real application. The performance is also mostly related to cost we incur to improve the performance. Key words: magnetic levitation, model reference adaptive control, PID, LQR