A huge potential to obtain clean energy exists from mixing water streams with different salt concentrations. Two membrane-based energy conversion techniques are evaluated: pressure-retarded osmosis and reverse electrodialysis. From the literature, a comparison is not possible since the reported performances are not comparable. A method was developed which allows for a comparison of both techniques at equal conditions, with respect to power density and energy recovery. Based on the results from the model calculations, each technique has its own field of application. Pressure-retarded osmosis seems to be more attractive for power generation using concentrated saline brines because of the higher power density combined with higher energy recovery. Reverse electrodialysis seems to be more attractive for power generation using seawater and river water. These conclusions are valid for present and latent performances of both techniques. According to the model, the potential performances of both techniques are much better than the current performances. In order to achieve these potential performances, the development of pressure-retarded osmosis must focus on membrane characteristics, i.e. increasing the water permeability of the membrane skin and optimization of the porous support. The development of reverse electrodialysis, however, must focus on system characteristics, i.e. optimization of the internal resistance, which is mainly determined by the width of the spacers.