A new method for the identification of the nonlinear Hammerstein model, consisting of a static nonlinear part in cascade with a linear dynamic part, is introduced. The static nonlinear part is modeled by a multilayer feedforward neural network (MFNN), and the linear part is modeled by an autoregressive moving average (ARMA) model. A recursive algorithm is developed for estimating the weights of the MFNN and the parameters of ARMA model. Simulation examples are included to illustrate the performance of the proposed method.