Nonaqueous redox flow batteries (RFBs) are one economically promising solution for meeting grid-scale energy storage needs at discharge durations of 10 h or more. However, membrane transport properties in nonaqueous systems are not as well understood as in water. Solvent-specific effects complicate efforts to understand transport in nonaqueous systems. Changing the solvent used to measure membrane transport properties causes changes in solvent uptake, which can mask other solvent-specific differences and trends. This study decoupled these effects by using crosslinked membranes with post-crosslinking solvent exchange steps to vary the membrane solvent uptake of three solvents that are suitable for RFBs. This approach enabled the independent study of solvent uptake and specific measurement solvent effects on membrane transport properties. The results revealed differences in polymer solvation between the measurement solvents, and these differences led to changes in the sensitivity of both ionic conductivity and uncharged active material permeability to solvent uptake. Additionally, these changes in sensitivity appeared to be independent of each other, e.g., a weak dependence of ionic conductivity on solvent uptake coupled with a strong dependence of permeability on solvent uptake was observed for some materials. As a result, the highest-performing membrane, a crosslinked phenoxyaniline trisulfonate-functionalized poly(phenylene oxide) membrane produced using acetonitrile as the de-swelling solvent and characterized using propylene carbonate, retained a high conductivity of 0.20 mS cm–1 while restricting active material permeability to less than 10–11 cm2 s–1. The reported solvent-specific behavior suggests that specific solvent–polymer interactions may provide a route to simultaneously increase ionic conductivity and decrease active material permeability, which would lead to more selective membranes to enable high-efficiency nonaqueous RFBs.