In order to improve the quality of the power injected into a grid, this paper presents a model predictive control strategy for three-phase four-leg grid-tied inverters. For the convenience of optimization, the discrete-time model of the inverter in which duty ratios are modeled as continuous control variables is investigated. A current tracking error oriented cost function is employed as a criterion to optimize duty ratios of the inverters. In order to eliminate the effects of sampling delay, a model predictive control with delay compensation method (MPC-DC) is proposed. Because there is a large amount of calculations in implementing predictive control algorithm, a double-CPU, namely FPGA plus DSP controller, is employed to implement parallel calculation, so as to reduce the computation time. Simulation and experimental results demonstrate the effectiveness of the proposed method.