The effect of dietary starch source and level on growth performance, feed utilization, apparent digestibility coefficients and liver enzyme activities involved in intermediary metabolism of gilthead sea bream juveniles was studied. Five isonitrogenous (47% crude protein) and isolipidic (15% crude lipids) diets were formulated to contain 10% native (diet NS10) or waxy (diet WS10) maize starch; 20% native (diet NS20) or waxy (diet WS20) maize starch or no starch (control). Diets were adjusted with a-cellulose. Another diet was formulated without carbohydrates, and contained 70% crude protein and 15% crude lipids (diet HP). Each diet was fed to triplicate groups of 30 fish (initial weight: 20 g) for 12 weeks. The HP group was fed to near satiation and the other 5 groups were fed on a pair-feeding scheme according to the group that ingested less feed (control diet group). The reduction of dietary protein level from 70% to 47% by the incorporation of 20% starch did not significantly affect gilthead sea bream growth performance or feed efficiency. Compared to the control diet, neither the level nor the nature of starch had any measurable effect on growth performance and feed efficiency. Digestibility of starch was unaffected by source or dietary inclusion level. Diet had no effect on plasma glucose levels, but liver glycogen was higher in diet groups NS20, WS20 and HP. Dietary carbohydrates increased GK and GOD enzyme activities and decreased ALAT and GDH enzyme activities while had only a minor effect on FBPase activity. The nature of dietary starch tested (native or waxy) had little influence on performance criteria. (C) 2007 Elsevier B.V. All rights reserved.