The present study was designed to investigate whether calorie source influences sodium and water metabolism and sympathetic activity during parenteral nutrition (PN). 20 New Zealand rabbits were starved until a mean weight loss of 18% was achieved and then re-fed for 6 days with 2 formulae of PN with different glucose-fat proportions. In the Glucose group (n = 9), 70% of non-protein calories were given as glucose while in the Lipid group (n = 11), 70% of non-protein calories were administered as lipids. Rabbits with a high glucose intake showed significantly higher weight gain (151 ± 87 vs. 52 ± 7 g, P = 0.01), water cumulative balance (542 ± 132 vs. 411 ± 87 ml; P = 0.02) and urinary metanephrine excretion (0.42 ± 0.12 vs. 0.30 ± 0.1 μmol/d, P = .03). Only in this group, urinary metanephrines correlated positively with water and sodium balances (r2 = 0.6; P = 0.02 and r2 = 0.7; P = 0.009 respectively). The Glucose group showed 2 different responses and in a second experiment 10 additional rabbits were added to this group to allow a statistical analysis of the response pattern: half of the animals increased their extracellular water (ECW) compartment while the remaining animals did not. The former group had higher sodium balance (13.9 ± 8 vs. 4.3 ± 5; P = 0.004) and wet lung weight (8.9 ± 0.9 vs. 7.9 ± 0.8; P < 0.05) after re-feeding and, at the beginning of PN, their serum aldosterone concentration were also higher (221 ± 11 vs. 130 ± 47 pmol/l; P < 0.05). In conclusion, glucose based PN appears to increase sympathetic activity and induce spurious weight gain due to markedly positive wate and sodium balances. Plasma aldosterone concentration at the end of starvation period influences sodium retention and ECW expansion during high glucose re-feeding. © 1994.