1. Two experiments are reported. In Expt 1, five male lambs of 26–33 kg were used to measure basal nitrogen excretion when the lambs were entirely sustained by an intraruminal infusion of 450 kJ/kg body-weight0·75 per d of volatile fatty acid (VFA) and were receiving no protein. In Expt 2, which was a conventional growth trial, the response to fish meal (66 or 132 g dry matter/d) of lambs given a control diet of sodium-hydroxide-treated barley straw was measured.

2. In Expt 1 the mean basal N excretion of the lambs was 429 (SE 21) mg N/kg body-weight0·75 per d. This exceeds current UK standards for the amino acid N of microbial origin which would be made available to the normally-fed host animal at a maintenance level of metabolizable energy intake.

3. In Expt 2 there was a clear growth response to the fish meal, which was greater (P < 0·05, single-tailed test) than that to be predicted from the energy content of the fish meal. There was no effect of fish meal on the voluntary intake of the basal diet, but there was a suggestion that the digestibility of the basal diet was improved.

4. It is concluded from Expt 1 that the basal requirement for amino acid N by lambs is three- to fourfold that currently recommended in the UK. This higher basal requirement should have resulted in a marked response to supplemental protein in Expt 2. The fact that the growth response in Expt 2 was less than anticipated may have been due to a combination of a slightly lower basal N excretion than that found in Expt 1, a higher yield of amino acids of microbial origin than current UK standards predict, and possibly to a change in the body composition of the lambs.