Statistically based experimental designs, based on the Plackett–Burman protocol, were applied to the optimization of biomass and arachidonic acid (ARA) production in Aureispira maritima shake-flask cultures. Tryptone and culture temperature were identified to have a significant effect on biomass production, whereas ARA production was only affected significantly by the pH and agitation rate. These four factors were subsequently optimized using response surface methodology. The validity of the optimum conditions was verified by separate experiments in which biomass and ARA yield were increased 4.02-fold (2.05 g l-1) and 3.59-fold (21.50 mg g-1), respectively, in 3-day fermentations. Under non-optimized culture conditions the corresponding values were 0.51 g l-1 and 5.99 mg g-1, respectively. The results suggest that A. maritima might be a potential strain for further large scale investigations to determine whether this bacterium