Critical evaluation of the nutrient requirements of an animal requires the use of a puri®ed reference diet. In this study, the eects of either a puri®ed reference experimental diet, a practical experimental diet and an imported commercial diet on the growth and nutritional condition of the prawn Penaeus monodon were compared. The reference diet, based on widely available ingredients of consistent and known nutrient content, supported excellent prawn growth when fed either ad libitum or subsatiation (» 75% of the mean initial satiation intake). In a 6-week experiment, growth (as percentage increase in weight, SEM) of the prawns fed ad libitum diered signi®cantly (P < 0.05) between each diet in the order: practical (142 10%), reference (121 9%) and commercial (91 16%). Growth for the reference diet fed subsatiation was 116 4%, and was not signi®cantly (P > 0.05) dierent from that obtained with the same diet fed ad libitum. Dry matter food conversion for the reference diet was signi®cantly (P < 0.05) better when fed subsatiation (1.58 0.08) rather than ad libitum (2.08 0.06) and better than that obtained with either the practical (3.40 0.15) or commercial (3.02 0.28) diets, which were not signi®cantly (P > 0.05) dierent from each other. The lipid content (mg g ±1 of prawn) of the digestive gland of prawns fed ad libitum was similar for reference, practical and commercial diets (6.03 0.38, 4.92 0.90 and 4.92 0.95 mg g ±1 of prawn, respectively), but signi®cantly (P < 0.05) higher than that for the reference diet fed subsatiation (4.73 0.38 mg g ±1 of prawn). KEY WORDS