Reverse osmosis is increasingly used in drinking water supply for treatment of fresh water sources which can directly result in high quality water. In practice reverse osmosis is never applied directly on fresh water sources, predominantly because of the occurrence of membrane fouling. Strategies to control membrane fouling in reverse osmosis without expensive and extensive pre-treatment include low flux operation, increase cross-flow velocities and improved (periodic) hydraulic cleaning. The objective of this research is to technically and economically compare minimal pre-treated reverse osmosis to state-of-the art ultrafiltration pre-treated reverse osmosis in surface water treatment to obtain high quality water. A parallel reverse osmosis system with six parallel 2540-type spiral wound membrane elements has been systematically operated during several months with only screens pre-treatment before reverse osmosis. A minimal pre-treated reverse osmosis system using only screens on surface water did not result in stable reverse osmosis operation, however, the ultrafiltration pre-treated reverse osmosis system also failed in the longer run. Lowering the membrane flux from 25 L/m2.h to 10 L/m2.h in minimal pre-treated RO resulted in an approximately 10-fold decrease of the membrane fouling rate, while not influencing the pressure drop increase. Periodical air water cleaning in minimal pre-treated RO resulted in an approximately 4-fold reduction of the spacer clogging rate, while not affecting the membrane fouling rate. The total cost of a minimal pre-treated low flux (10 L/m2.h) operated RO was significantly lower (0.46 €/m3) compared to state-of-the-art UF pre-treated RO at normal flux (25 L/m2.h) (1.21 €/m3). The cost difference in financial space provides an opportunity to develop minimal pre-treated RO systems with improved control of membrane fouling and spacer clogging.