This study evaluates two different strategies for arsenic removal in a direct filtration application at the Los Angeles Aqueduct Filtration Plant (LAAFP): use of elevated ferric iron coagulant doses, and use of alternate types of filter media. Both strategies exhibit some potential for improved arsenic removal (from approximately 30% to approximately 60-70%), but also have limited application in the direct filtration process. Within the framework of the existing direct filtration process, a significant improvement in arsenic removal can be attained through small increases in the ferric iron coagulant dose, but if the iron dose is raised too high, the performance of filters is compromised due to premature turbidity breakthrough. Turbidity breakthrough can be delayed by decreasing the filtration rate, by optimizing the cationic polymer dose, and by using an anionic polymer filter aid, but severe breakthrough still occurs at very high iron doses. Operationally, at this location, the benefits of improved arsenic removal must be weighed against costs and against other treatment objectives. The use of alternative filter medias, such as iron coated anthracite, also give significantly improved arsenic removal, especially at low ferric iron coagulant doses. Questions still remain regarding media longevity and economic feasibility.