A high-rate algae pond (HRAP) was tested as a second loop of water treatment in a recirculating fish rearing system to reduce water requirements and nutrient discharge levels. Three duplicated groups of sea bass (mean initial body weight 3 5 I I g) were reared under different system conditions (flow-through system, recirculating system and recirculating system with HRAP) for I year. Fish survival rate was higher in the system with HRAP, and their mean body weight was statistically higher (p<0.05) during the month of maximal climatic conditions for algae photosynthesis. After 266 days of experiment, mean fish weight was significantly higher (p<0.05) in the flow-through system than in the recirculating systems. Final fish weight was 15% higher in the flow-through system than in recirculating systems and was related to a higher daily-ingested food. The comparison of mean annual nutrient concentrations in the recirculating systems gives a statistically significant reduction of 25% of nitrogen (p<0.01) and 9% of phosphates (p<0.01) due to the HRAP. Absorption of nitrate form is responsible for nitrogen removal and is related to climatic conditions for algae growth. The phosphate precipitation at high pH (above 8) was not considered. The maximal removal rates were 0.5+/-0.2 g N m(-2) day(-1) and 0.03+/-0.02 g P m(-2) day(-1) for nitrates and phosphates, respectively, and were obtained during the optimal climatic conditions and the shortest algae harvesting frequency. These results are favourable to complete reuse of the HRAP treated water, all year long, in the recirculating rearing systems. In order to improve nitrate and phosphate removal rates, a periodic harvesting of algae is necessary and a higher water inflow in HRAP should partly make up for inorganic carbon depletion in high nitrate and phosphate effluents.