A pilot study was conducted from October 2007 to November 2010 to establish the long-term feasibility of using reverse osmosis (RO) treatment to manage salt levels in Central Arizona Project water. Pretreatments consisting of microfiltration (MF) and slow sand filtration (SSF) were compared based on performance-turbidity removal, silt density index (SDI), volume treated between cleaning events and protection of downstream RO-during side-by-side operation over a yearlong period. SSF always produced feed water that was suitable for RO treatment (SDI < 5). However, MF consistently provided filtrate with SDI < 3, and long-term RO performance improved significantly with MF as pretreatment. Although the economic costs of MF and SSF pretreatments are similar; MF is preferred based on the quality of treated water and stability of downstream RO operation. (C) 2013 Elsevier B.V. All rights reserved.