Abstract: By transforming the macroscopic traffic flow model into a more general discrete-time nonlinear system model, an iterative learning identification method is developed to estimate the parameters of the more general discrete-time nonlinear system, so the macroscopic traffic flow parameters as well, based on the repeatability of the macroscopic traffic flow behavior in a freeway. With rigorous analysis, it is shown that the proposed learning identification scheme can guarantee the convergence and robustness. A number of simulation results are provided to demonstrate the efficacy of the proposed approach.