H 2 O Gluconeogenesis Glycogenolysis Glucose 6-phosphatase Glucokinase NMR Sources of blood glucose in European seabass (initial weight 218.0 ± 43.0 g; mean ± S.D., n = 18) were quantified by supplementing seawater with deuterated water (5%-2 H 2 O) for 72 h and analyzing blood glucose 2 H-enrichments by 2 H NMR. Three different nutritional states were studied: continuously fed, 21-day of fast and 21-day fast followed by 3 days of refeeding. Plasma glucose levels (mM) were 10.7 ± 6.3 (fed), 4.8 ± 1.2 (fasted), and 9.3 ± 1.4 (refed) (means ± S.D., n = 6), showing poor glycemic control. For all conditions, 2 H-enrichment of glucose position 5 was equivalent to that of position 2 indicating that blood glucose appearance from endogenous glucose 6-phosphate (G6P) was derived by gluconeogenesis. G6P-derived glucose accounted for 65 ± 7% and 44 ± 10% of blood glucose appearance in fed and refed fish, respectively, with the unlabeled fraction assumed to be derived from dietary carbohydrate (35 ± 7% and 56 ± 10%, respectively). For 21-day fasted fish, blood glucose appearance also had significant contributions from unlabeled glucose (52 ± 16%) despite the unavailability of dietary carbohydrates. To assess the role of hepatic enzymes in glycemic control, activity and mRNA levels of hepatic glucokinase (GK) and glucose 6-phosphatase (G6Pase) were assessed. Both G6Pase activity and expression declined with fasting indicating the absence of a classical counter-regulatory stimulation of hepatic glucose production in response to declining glucose levels. GK activities were basal during fed and fasted conditions, but were strongly stimulated by refeeding. Overall, hepatic G6Pase and GK showed limited capacity in regulating glucose levels between feeding and fasting states.