A reverse osmosis study with a threefold objective was conducted. Tubular membranes prepared from trans esterified cellulose acetate were compared with membranes made from commercially available cellulose acetate (control). Next an evaluation was conducted of the in-situ regenerable membrane reverse osmosis design on wastewater. Finally the membranes were subjected to carbon treated secondary effluents, primary effluents, and concentrated primary effluents. Results showed that the modified tubular membranes produced fluxes slightly greater than the control membranes, while in-situ regenerable membranes produced fluxes significantly below those obtained with tubular units. On carbon treated secondary effluents, modified tubular membranes produced an overall average product water flux of 15 gfd, compared to 10 gfd for the control. Initial levels were 15 and 25 gfd for the control and the modified membranes respectively, and these fluxes could be maintained by cleaning every 10 days with an enzyme presoak. Flux for the concentrated primary effluent (9 times the solids in primary effluent) stabilized at 4 to 5 gfd. Removal of wastewater constituents remained nearly constant at 90 to 100% for all three feeds, thus establishing the technical feasibility.