All input experimental errors for membrane flux determination were analyzed. The overall error in pressure did not largely influence the determination of membrane flux. The main contribution to errors was ascribed to flow meters and membrane thickness. The choice of an optimal operational range for input parameters led to elimination of at least 2/3 of the theoretical relative error for subsequent evaluation of the permeability of individual species. This is especially valuable for low-permeable membranes, whereas a very low relative theoretical error can be obtained in the case of highly permeable materials.