This paper presents a discussion of two main approaches for selecting a nanofiltration (NF) or reverse osmosis (RO) technology and a method for cost comparison over a full life cycle. The first approach is to use NF membranes, which operate at low feed pressure, but provide less contaminant rejection. Using this approach, the degree to which blending can be used is limited and overall horsepower required for treatment remains high. The second approach is to use RO membranes, which operate at high pressure, but provide high contaminant rejection, and therefore reduce the volume of treated water required to produce a specified blended finished water quality. This paper shows that RO in certain cases may be more cost-effective than NF. This is due in part to the fact that RO product waters in many cases may be blended to a greater extent than NF product water. Greater blending may also allow for a reduction in capital and subsequent O&M expenditures. This paper presents a study of the impacts of these factors on full life cycle costs. Considerations are given to capital costs associated with a new membrane plant, feed water quality, finished water quality goals, membrane technology (NF vs. RO), pressure energy requirements, rejection, the degree of blending, and the amount of system recovery that can be achieved. Includes tables, figures.