In early January of 1996, Dominguez Water Corporation requested that Flow Tech Industries Inc. develop a plan to treat a municipal well that had exhibited fluctuating levels of fluoride concentration ranging from 3.6-5.3 ppm. The well serves a small community located near a large lake. The lake is a popular recreational area serving Bakersfield and Los Angeles, California. The goal of the project was to treat the raw water sufficient to meet current and anticipated California Maximum Contaminant Levels for drinking water. The cost of the water produced by the system had to be a competitive alternative to developing new wells and/or tying into other municipal supply Lines. Development of the project from concept to pilot study, to full-scale plant operation, and first year operating costs will be discussed. The paper focuses on three major issues: 1) the value and limitations of using pilot data from small membrane systems to design and build full-scale membrane plants; 2) specific machine design features that allow full-scale plants the flexibility to operate outside of the initial design parameters. Designing in features that will accommodate a wide range of flow and contaminant levels must be balanced against the capital costs incurred by utilizing such design features; 3) operation and maintenance of the existing plant including first year operating costs.