Riverbank filtration (RBF) is a natural process, using alluvial aquifers to remove contaminants and pathogens in river water for the production of drinking water as a low-cost water treatment technology. This study illustrates the development and application of feed-forward backpropagation network (BPN) as a type of artificial neural networks. The BPN prediction results produced good agreement with measured data at a correlation coefficient above 0.98 for filtrate water quality parameters, including temperature as well as turbidity, heterotrophic bacteria, and coliform removal.