The novel exopolysaccharide bioflocculant HBF-3 is produced by Halomonas sp. V3a', which is a mutant strain of the deep-sea bacterium Halomonas sp. V3a. Response surface methodology (RSM) was employed to optimize the production medium for increasing HBF-3 production. Using a Plackett-Burman experimental design to aid in the first step of optimization, edible glucose, MgSO(4)center dot 7H(2)O, and NH(4)Cl were found to be significant factors affecting HBF-3 production. To determine the optimal concentration of each significant variable, a central composite design was employed. Based on response surface and canonical analysis, the optimum concentrations of the critical components were obtained as follows: edible glucose, 16.14 g/l; MgSO(4)center dot 7H(2)O, 2.73 g/l; and NH(4)Cl, 1.97 g/l. HBF-3 production obtained by using the optimized medium was 4.52 g/l, which was in close agreement with the predicted value of 4.55 g/l. By scaling up fermentation from flask to fermenter, HBF-3 production was further increased to 5.58 g/l. (C) 2009 Elsevier Ltd. All rights reserved.