A study of the fractionation of nitrogen isotopes in the diet by cattle is described and the results discussed. Compared with the diet, urine had a lower ratio of ¹⁵N to ¹⁴N, but faeces, blood and milk all had a higher ratio. It is argued that the use of natural ¹⁵N as a tracer in grazed ecosystems is more complicated than was at first thought.