Abstract: Present adaptive predistortion structures make against the utilization of efficient least-square algorithms in the parameters update of Hammerstein predistorter. In order to solve the problem, a novel adaptive structure is proposed. Using the structure, errors of the two subsystems of a Hammerstein predistorter can be obtained, so efficient least-square algorithm can be used in the parameters update of a Hammerstein predistorter directly. By this means, the effect on the performance of predistorter induced by the structure error and the imprecision of subsystem errors become evitable. It is confirmed by computer simulation that Hammerstein predistorter could more efficiently compensate the nonlinear distortion of power amplifier with memory effects by the proposed adaptive predistortion structure.