Fifteen sheep (53 kg), with catheters in a hepatic vein, the portal vein and a mesenteric vein and artery, were offered a bromegrass (Bromus inermis) hay (104 g crude protein (CP), 700 g neutral-detergent fibre and 65 g acid-detergent lignin/kg dry matter (DM)) ad lib. with different supplements to determine the effects on net flux of oxygen and nutrients across the portal-drained viscera (PDV) and liver. The sheep were unsupplemented (Control) or received 5 g DM/kg body weight (BW) of ground maize (M), 7 g DM/kg BW of soya-bean hulls (H) or 0.73 g DM/kg BW of a mix of feedstuffs high in rumen-undegradable protein (P). Apparent digestible energy (DE) intakes were 5·3, 10·4, 10·6 and 6·7 (SE 0·74) MJ/d and apparent digestible CP intakes were 37, 50, 79 and 68 (SE 4·3) g/d for Control, M, H and P treatments respectively. Splanchnic tissue oxygen consumption rates were 0·23, 0·32, 0·30 and 0·27 (SE 0·054) mol/h, and oxidative metabolism accounted for 0·46, 0·31, 0·33 and 0·47 (SE 0·051) of DE intakes for Control, M, H and P treatments respectively. Supplements increased (P 0·05) release of α-amino nitrogen (AAN) by the PDV (4·2, 17·5, 19·6 and 18·1 mmol/h for Control, M, H and P treatments respectively). Splanchnic net flux of AAN was not affected by supplement treatments. Hepatic release of urea-N was increased (P 0·05) by supplement treatments (27, 40, 46 and 44 mmol/h for Control, M, H and P respectively); the P treatment increased (P 0·05) and the H treatment tended (P = 0·10) to increase splanchnic release of urea-N (7, 10, 20 and 27 mmol/h for Control, M, H and P treatments respectively). Net flux of glucose across the PDV was -4·6, 1·4, -5·6 and -7·2 (SE 1·65) mmol/h for Control, M, H and P treatments respectively. Hepatic glucose released averaged 23 (SE 2·0) mmol/h and was not affected by treatment. Treatments M and H increased (P 0·05) PDV release of propionate compared with the Control treatment (4·5, 15·5, 16·8 and 7·7 mmol/h for Control, M, H and P treatments respectively). Release of acetate by the PDV was 43, 97, 118 and 67 (SE 23·9) mmol/h for Control, M, H and P treatments respectively. In summary, different supplements of low-quality grass did not increase the efficiency of N metabolism by splanchnic tissues. Treatment P had little effect on net flux across splanchnic tissues of glucose, L-lactate. β-hydroxybutyrate and volatile fatty acids (VFA). Overall, treatments M and H had similar effects on splanchnic net fluxes of VFA and L-lactate whereas butyrate and β-hydroxybutyrate releases by the PDV were increased by treatment M.