Three isonitrogenous (48% protein) and isolipidic (12% lipids) diets were formulated to contain 25% raw starch (RS), 25% gelatinized starch (GS) or 12.5% raw+12.5% gelatinized starch (RS/GS). Another diet was formulated with no carbohydrates (HP) containing 63% protein and 18% lipid. Each experimental diet was fed to triplicate groups of 25 fish (IBW=6 g) for 10 weeks.The reduction of dietary protein level from 63% to 48%, by starch incorporation, did not affect growth rate or protein and energy retention. For the diets containing starch, specific growth rate of fish fed the GS diet and feed efficiency for the RS diet were significantly lower than values obtained for the other diets. Protein digestibility was independent of diet composition. Energy and starch digestibility significantly increased with the increase in dietary gelatinized starch.At the end of the growth trial there were no significant differences in whole body composition between groups. Hepatosomatic and visceral indices of fish fed the RS and HP diets were significantly lower and the hepatic lipid content of fish fed the GS diet was significantly higher than in the other groups. Muscle lipid content was unaffected by diet composition. After 16 hours of fasting, the plasma glucose level was identical for all dietary treatments.Nitrogen (g kg−1 day−1) and energy retention (kJ kg−1 day−1) were unaffected by dietary treatment. However, a trend towards increased nitrogen and energy retention (% N or E intake) was observed with the increase of dietary gelatinized starch. Ammonia excretion (mg kg−1 day−1) and heat increment (kJ kg−1 day−1) were significantly lower for sea bass fed the GS diet.In conclusion, the total replacement of raw starch by gelatinized starch significantly reduced growth and feed intake, while improving feed efficiency. Overall, the best results were achieved with the diet RS/GS.