This paper presents a mathematical framework that can be used to determine the flow distributions for a step-feed activated sludge process that result in maximum nitrogen removal. The model indicates that nitrogen removal efficiency in a step-feed activated sludge process is highly dependent on the ultimate biochemical oxygen demand (BOD L )-to-total Kjeldahl nitrogen (TKN) ratio of the wastewater. For typical domestic wastewater, which has a relatively high BOD L -to-TKN ratio, the step-feed process will outperform the Modified Ludzack-Ettinger process for nitrogen removal, when the flow to each step is optimally distributed. Using plantspecific water quality data and operating conditions from a 1-year period, nitrogen removal performance for four step-feed activated sludge plants operated by the Sanitation Districts of Los Angeles County (California) was calculated using the developed model. The calculated nitrogen removal efficiencies match well with the actual plant performance data. These results validate the model as a useful tool for predicting nitrogen removal in a stepfeed activated sludge process. Other analyses revealed that improvements in nitrogen removal at existing facilities are achievable by adjusting the split of primary effluent flow to each anoxic zone several times during the day. The timing of the adjustments and the optimal flow splits can be determined from data on diurnal fluctuations in BOD L and TKN concentrations. An example is provided to illustrate the application of such an operating strategy and the potential enhancement of nitrogen removal. Water Environ. Res., 79, 367 (2007).