In this paper, an online BP neural network (BPNN) compensate control scheme based on inverse system method is presented for a class of single-input-single-output nonlinear systems. Firstly, the error between the α-th derivative of the system output and the pseudo-control is analyzed and a BPNN is designed to compensate the error. Then, an adaptive algorithm of the BPNN, designed based on the Lyapunov stability theory, proves that tracking error of closed-loop system and weight estimation error of BPNN are uniform ultimate boundedness. Simulations for three nonlinear systems demonstrate the validity of the proposed control scheme.