Reverse osmosis (RO) experiments are conducted in the laboratory using a cellulose acetate membrane in a flat disk cell. The data are used to estimate membrane parameters and mass transfer coefficient using Kimura-Sourirajan analysis (KSA) and by a new graphical method (GM). Even though the origin of the two methods is similar, the membrane parameters and coefficients are calculated using different procedures. The parameters estimated from the KSA method, in which every parameter is estimated at each data point, are prone to experimental errors and show marked variation with operating conditions. In contrast, the graphical method, in which data at different pressures but constant feed flow rate and constant feed concentration are used in a simple graphical procedure, show that the estimated membrane parameters are reasonably constant. It is therefore shown that the estimation of parameters using the KSA method may lead to the conclusion that the membrane parameters are functions of operating conditions such as pressure whereas in reality the parameters may not be functions of operating conditions at all.