The design and operation of nanofiltration (NF) processes requires the use of good predictive methods such as those based on the extended Nernst-Planck equation. However, use of such methods generally requires a prior characterization of membrane properties which depends on extensive experimentation. The present paper develops an approximate characterization method which uses specification data obtained from the membrane manufacturers. For one membrane, NF-CA30 from Hoechst, the characterization parameters obtained using this approximate method have been compared to those obtained using extensive experimental data. Very good agreement was found, confirming the validity of the approximate method. The method was subsequently used to characterize 29 NF membranes from 10 different manufacturers. The results provide a quantification of the ranges of the three key parameters (effective pore radius (r(p)), ratio of effective membrane charge to solution membrane thickness to membrane porosity (Delta x/A(k))) occurring in commercial NF membranes. This simplified characterization method, when used in conjunction with a predictive model based, for example, on the extended Nernst-Planck equation, should be very useful in the selection of membranes suitable for specific applications. Further, knowing the range of parameters for NF membranes available in the market, the contribution of each transport mechanism in the membrane (diffusion, electromigration, and convection) has been assessed. The analysis shows that generally all three transport mechanisms are important. It would be unwise to exclude any of the mechanisms from a theoretical description of nanofiltration which aims to have wide applicability.