Pressure retarded osmosis (PRO) and reverse electrodialysis (RED) are emerging membrane-based technologies that can convert chemical energy in salinity gradients to useful work. The two processes have intrinsically different working principles: controlled mixing in PRO is achieved by water permeation across salt-rejecting membranes, whereas RED is driven by ion flux across charged membranes. This study compares the energy efficiency and power density performance of PRO and RED with simulated technologically available membranes for natural, anthropogenic, and engineered salinity gradients (seawater–river water, desalination brine–wastewater, and synthetic hypersaline solutions, respectively). The analysis shows that PRO can achieve both greater efficiencies (54–56%) and higher power densities (2.4–38 W/m2) than RED (18–38% and 0.77–1.2 W/m2). The superior efficiency is attributed to the ability of PRO membranes to more effectively utilize the salinity difference to drive water permeation and better suppress the detrimental leakage of salts. On the other hand, the low conductivity of currently available ion exchange membranes impedes RED ion flux and, thus, constrains the power density. Both technologies exhibit a trade-off between efficiency and power density: employing more permeable but less selective membranes can enhance the power density, but undesired entropy production due to uncontrolled mixing increases and some efficiency is sacrificed. When the concentration difference is increased (i.e., natural → anthropogenic → engineered salinity gradients), PRO osmotic pressure difference rises proportionally but not so for RED Nernst potential, which has logarithmic dependence on the solution concentration. Because of this inherently different characteristic, RED is unable to take advantage of larger salinity gradients, whereas PRO power density is considerably enhanced. Additionally, high solution concentrations suppress the Donnan exclusion effect of the charged RED membranes, severely reducing the permselectivity and diminishing the energy conversion efficiency. This study indicates that PRO is more suitable to extract energy from a range of salinity gradients, while significant advancements in ion exchange membranes are likely necessary for RED to be competitive with PRO.