For the characteristics of wave-piercing catamaran longitudinal motion when sailing with high speed in random waves, this paper proposes a state estimation algorithm for wave-piercing catamaran longitudinal motion based on EKF. On the one hand noise is filtered out from the noise pollution measurement signal as much as possible, and the desired signal is separated therefrom; on the other hand, all state information of the system can be obtained by simply selecting the pitch angle and heave displacement of wave-piercing catamaran. Theoretical analysis and simulation results show that the proposed state estimation algorithm can achieve satisfactory filtering accuracy and filtering efficiency.