1. In a comparative-slaughter experiment, individually rationed wether lambs initially of 42 kg were given 235, 362 or 456 kJ metabolizable energy (ME)/kg live weight (LW)0'75 per d as sodium hydroxide-treated barley straw with urea (six lambs per treatment), or NaOH-treated barley straw with urea plus 125 g/d white-fish meal to give 307 or 488 kJ ME/kg LW0"75 per d (seven lambs per treatment) for 92 d.2. All unsupplemented lambs lost both fat and body protein. The changes in fat were -3.53, -2.75 and -1.40 (SE 0.59) kg (initial value 8.6 kg), and the changes in body protein were -0.47, -0.09 and -0.14 (SE 0.13) kg (initial value 4.9 kg) for the three unsupplemented groups respectively. When supplemented with fish meal, fat was again lost as -1.53 and -0.93 (SE 0.55) kg, but wool-free body protein was increased, and gains were 0.48 and 0.89 (SE 0.12) kg for the two supplemented groups respectively. All animals lost wool-free body energy, total changes being-150, -11 1, -59 and -49 and -16 MJ respectively. When corrected to an equal ME intake the supplemented lambs, when compared with the unsupplemented lambs, gained (instead of losing) body protein (P < 0.001) and lost less fat (P < 0.05). Wool growth did not respond to supplemental protein, but was related to ME intake with an increase of 0-78 g wool fibre for each additional MJ ME.3. The maintenance requirements of the unsupplemented and supplemented groups respectively were estimated by regression analysis to be 554 and 496 kJ ME/kg LW0'76 per d. The apparent utilization of ME below energy equilibrium (k,) was 0.31 (SE 0.08) for the unsupplemented animals, and 0.12 (SE 0.10) for the supplemented animals, well below a k , of 0.70 which current UK standards (Agricultural Research Council, 1980) would predict. Most of these differences could be reconciled if basal metabolism was assumed not to be constant. 4.It is concluded that lambs in negative energy balance can continue lean body growth at the expense of body fat, provided sufficient dietary protein is available. It is also concluded that since the animals at the lowest ME intakes required less ME than predicted by current feeding standards, the effect was that it would have been difficult to distinguish between the apparent utilization of ME for maintenance (k,) and for fattening (kf).That the nitrogen requirements of ruminants are a combination of the needs of the host animal and of the rumen micro-organisms has long been recognized. However, only recently have there been attempts to describe ruminant protein requirements as an integration of these separate components (Institut National de la Recherche Agronomique (INRA), 1978 ; Agricultural Research Council (ARC), 1980), for the complex and inter-related activities of the micro-organisms and host are difficult to measure.The development of the total-infusion technique with which ruminants are entirely maintained by the intragastric infusion of volatile fatty acids and protein (0rskov et al. 1979) has provided the means whereby the N metabolism of the host...