In a reverse electro dialysis (RED) installation, power is produced from the chemical potential difference between salt- and freshwater using ion-selective membranes. In order to make a RED plant commercially feasible, large amounts of salt- and freshwater are needed. At the Afsluitdijk, salt water can be extracted from the Wadden Sea and freshwater from Lake IJssel. The water from the Wadden Sea, however, contains high concentrations of suspended particles (on average ca 50 mg l-1). These particles adversely affect the efficiency of the plant and need to be removed from the water before it enters the membrane stacks. Shellfish are efficient filterfeeders that are capable to filter large amounts of suspended solids from the water. Therefore, it has been suggested that shellfish can be used to pre-filter the water from the Wadden Sea, before it enters the reverse electro dialysis power plant. In a previous model study, it has been shown that, depending on residence times and amount of shellfish, mussels are capable to remove 50% of the suspended particles from the water. In this report, the results of a large-scale experimental study, that was executed to test if shellfish can be used as pre-filter for marine intake water, are presented. Two consecutive experiments (in spring and in summer) were run at the test facility at the Afsluitdijk using the blue mussel (Mytilus edulis). Filtration efficiency of marine intake water by shellfish was measured in a flow-through system containing mussels and compared to a control flow-through system without shellfish. The flow-through system was designed to create low flow velocities allowing the larger suspended particles (faeces and pseudofaeces) to sink and accumulate at the cone-shaped bottom of the tank. The accumulated deposits can be quantified and removed from the tank. The results of the large-scale experiments showed that depending on the set-up, the mussels were able to remove 6-11% of the incoming sediment over a period of 2 to 3 months. Within this period, moments occurred where more than 50% of the suspended particles were removed by the mussels. During the experiment in spring, the deposition rate in the tank with mussels (average 95 kg fresh weight) was on average 2.9 kg day-1 while the deposition rate in the control tank, without mussels, was 1.2 kg day-1. During the experiment in summer, the deposition rates in the mussel (average 35 kg fresh weight) and control tanks were 2.4 and 0.4 kg day-1, respectively. At a flow rate of 5 m3 per hour 11% of the incoming suspended matter was filtered by on average 35 kg of mussels. A powerplant with a capacity of 10 MW needs 10 m3 s-1 sea water (36 000 m3 per hour). To pre-filter the water with an efficiency of 11%, a total of 252 000 kg of mussels are needed, producing a total of about 16 tons of biodeposits per day. For upscaling purposes the design of the shellfish filtration system should be optimised to increase filtration efficiency of the mussels, minimize resuspension of (pseudo)faeces and increase the efficiency to remove the produced (pseudo)faeces from the systems. During the experiments, the mussels survived and even increased in weight. In the spring experiment the mussels grew on average 5 mm in length, increased their fresh weight (shell + tissue) by 62% and increased their flesh percentage on average from 12.7% to 20.6% over the course of the experiment executed between March and May. Approximately 57% of the mussels survived the experiment. Over the course of the experiment executed in summer (June – September), mussels grew on average 3 mm in length and increased their fresh weight (shell + tissue) by 15%. Percentage flesh decreased from 12.7% to 8.0% over the course of the experiment. The survival of the mussels during the second experiment was lower (38%) than in the first experiment. It is expected that modifications in the design of the set-up will increase the efficiency of the sediment removal from the water and the survival of the mussels.