This paper presents a novel approach for rail pressure control of GDI engines. The control objective is to make the rail pressure track the given reference. Different from the existing nonlinear controller design methods, the proposed procedure can be divided into 3 steps: 1) A steady state control is deduced, playing the similar role of the mapbased control strategy that widely used in automotive control; 2) A feedforward control is derived based on the reference dynamics; 3) A feedback control is also considered which can be rearranged into a state-dependent PID. The proposed method provides a concise derivation procedure and the structure is comparable to those widely used in modern automotive control. Finally, simulation results in the environment of commercial software AMESim with actual system parameters show the efficiency of the proposed approach.