When the received data contains the desired signal, the performance of most mainlobe interference suppression methods is seriously degraded. To solve this problem, a novel mainlobe interference suppression method based on eigen-projection processing and covariance matrix reconstruction for array antenna is proposed in this paper. The method constructs the sidelobe interference plus noise covariance matrix (SINCM) and remove the mainlobe data of the covariance matrix. Then, it uses eigen-projection technology to remove the mainlobe interference in the received data and correct the SINCM to make the proposed method has better ability to suppress the mainlobe interference and sidelobe interference when the received data contains the desired signal. Simulation results demonstrate the superior performance of the proposed beamformer relative to other existing methods.