A new concept for reverse osmosis is identified based on the use of a temperature gradient instead of pressure. When the temperature of the permeate-side of the membrane is higher than the feed-side then a significant driving force exists for water transport, which can overcome the osmotic pressure. The thermodynamics for this approach are developed within the paper, and as a result we have developed a single expression for driving force across a membrane for variable temperature, pressure and concentration. The thermodynamic predictions suggest for seawater a temperature difference of less than 1 o C is needed to overcome the osmotic pressure, and less than 3 o C to sustain a water flux equivalent to current reverse osmosis processes. Experimental investigation confirmed the temperature-dependence of water flux and the ability to carry out reverse osmosis at atmospheric pressure. The effect of temperature gradient and salinity on water flux was tested at ambient pressures and found to be in good agreement with the manufacturer-quoted permeability. The concept identified in this work has the potential to allow reverse osmosis to be carried out without the need for costly high pressure pumps and energy recovery systems, with energy requirements predicted to be lower than 2.0 kWh/m 3 .