Forward osmosis (FO) is an emerging novel filtration technology used to reclaim water from waste streams or concentrate the valuable component. Accurate water flux prediction of FO is essential to facilitate process optimization and scale-up of membrane systems The overall objective of this investigation was to propose and confirm a model for prediction of the structural parameter (S) which is required for flux determination of FO. Experimental measurements of water permeability (A) and solute permeability (B) revealed that these magnitudes were different for FO mode (Active layer facing feed) compared to PRO (Support layer facing feed) mode using the non-pressurized method. A comprehensive review of research literature identified eleven (11) models for prediction of tortuosity. After measurement of the membrane porosity, the structural parameter (S) was predicted for each of the eleven tortuosity models. Experimental validation was conducted to statistically determine the accuracy and bias of the proposed models. Statistical results of root mean square error, accuracy factor, and bias factor exhibited that the newly proposed equation originated from fractal theory is most accurate for water flux predictions in both FO and PRO mode of the two commercial membranes. Our results indicate that proper A and B values should be obtained and used depending on the operation mode while S can be mathematically predicted based on fractal theory.