The magnetic levitation system (Maglev) is highly nonlinear and unstable system. The Stabilization of magnetic levitation system is quite difficult without controller implication. But, the control effort itself has own dynamics that increases the overall dynamics of the system. For optimal control, one can need the entire states of the system should be measurable which requires a costly reconstructive filter. So that, a reduced order modeling based suboptimal controller is proposed. The model reduction technique provides a platform to easy understanding of original system. It also provides a simpler dynamics for easy handling and control aspect. The suboptimal controller is cost effective and hardware requirement is substantially decreased. The simulation results verify the effectiveness of proposed method. In addition, the advantage of proposed controller is indicated in comparison with an optimal control method.