In theory, the batch reverse osmosis (RO) process achieves the lowest practical energy consumption by varying pressure over time. However, few batch RO systems have been built and operated. We have designed, built, and operated the first "true" batch RO prototype using a flexible bladder. The flexible bladder serves as the high-pressure variable-volume tank that is inherent to true batch RO designs (as opposed to batch RO with energy recovery devices). We experimentally validated a model of batch RO energy consumption (≤ 2.7% difference) by measuring the hydraulic work of the high pressure and circulation pumps. We find that batch RO energy consumption will be greater than expected mostly due to salt retention, a problem neglected by most previous studies. However, despite operating at elevated salinity and flux conditions, batch RO can still save energy relative to single-stage and multi-stage continuous systems. For a seawater desalination plant (35 g/kg intake, 50% recovery, 15 L m −2 h −1), our newly-validated model predicts that batch RO would save 11% energy compared to a single-stage continuous RO plant. Our work demonstrates that batch RO is an energy-efficient process with the potential to reduce the cost of water desalination.