Pilot-scale studies were conducted to evaluate the removal of Cryptosporidium oocysts by direct filtration processes. Key water treatment parameters examined during this study included optimal metal coagulant/organic polymer combinations, pre-oxidation with either chlorine or ozone, coagulation pH depression and dual- and tri-media filtration at various filtration rates. Studies were initially conducted to optimize direct filtration for the removal of turbidity, particles, and aerobic spore-forming bacteria. Filter challenges were then conducted using seeded Cryptosporidium oocysts under optimized conditions. Under these test filtration conditions, >3 log seeded oocyst removals were observed with both alum and ferric chloride treated water. Greater removals of turbidity, particles, aerobic spores and Cryptosporidium oocysts could be achieved with ferric chloride under optimized conditions. However, the use of ferric chloride resulted in much greater rates of head loss accumulation and necessitated more frequent backwashing. As a result, filter effluent turbidity and particle levels, when averaged over the length of the filter run, were similar for alum and ferric chloride treated water.