An application of the extended Kalman filter for a nonlinear model of an activated sludge process (ASP), working in an alternating aerobic-anoxic phase medium, is proposed. The filter is used to estimate both the states and non-stationary disturbances, to better evaluate changes in operating conditions. It is to be pointed out that, according to the structure of the reduced nonlinear system, the observer is a two-alternated-observer scheme. Filtering results are shown by using both experimental data and simulations.