The City of Winnipeg (pop. 630,000) obtains its drinking water from Shoal Lake, which borders the Canadian provinces of Manitoba and Ontario, and the state of Minnesota, USA. The raw water supply is chlorinated at the intake for pathogen, taste and odor, slime, and zebra mussel control. The water then flows through a 159-km long aqueduct to the Deacon Reservoir in Winnipeg. Water for distribution is rechlorinated at the outlet of the reservoir, sent to two covered distribution system reservoirs, and then delivered to consumers. The City completed an intensive Phase 2 pilot water treatment program in 1997. Phase 2 tested the most suitable treatment processes for Winnipeg's water quality, resulting in a recommendation for a baseline water treatment process, which included conventional dissolved air flotation (DAF). In November 2002, a Value Engineering review team identified high-rate DAF as a potential cost savings alternative for consideration by the City. In 2003, the City contracted a consultant team from Earth Tech and CH2M HILL to determine the suitability and limits of the high-rate DAF process. The pilot system consisted of ferric chloride coagulation, high-rate DAF, intermediate ozonation, and deep bed filtration, and was evaluated under cool, cold, and warm water conditions. The high-rate DAF pilot tests were conducted in 2003 and the suitability of high-rate DAF was determined by comparing the process performance with data from the conventional DAF process tested in 1997. The key parameters investigated were the DAF loading rates, post-filter turbidity and particle counts, total organic carbon (TOC) removal, and unit filter run volumes. The potential for the freeze/thaw dewatering of plant residuals was also examined. High-rate DAF testing with and without lamella plates installed in the DAF tank was conducted during the warm and cold water test periods to determine their effect on the DAF loading rate. Without the lamella plates installed, successful test results were achieved at DAF loading rates of 45 and 40 m/h in the warm and cold water test periods, respectively. With the lamella plates installed, DAF loading rates as high as 50 and 45 m/h were reached in warm and cold water conditions, respectively. In 1997, the conventional DAF process was limited to loading rates of 20 and 15 m/h in the warm and cold water test periods. During the high-rate DAF pilot trials, filter run volumes were as high as 900 m3/m2, and post-DAF and filter turbidities were consistently less than 1.0 and 0.1 NTU, respectively. On average, filtered water particles were less than 10 /mL (total particles>2 um) and the TOC reduction through the high-rate DAF was between 60 and 70%. Based on the success of the high-rate DAF pilot trials, conventional DAF was replaced with high-rate DAF as part of the final water treatment process. Includes tables, figures.