Tracers have been extensively applied in lactating animals, for reasons that are not hard to determine; it is easy to obtain samples of milk and, in dairy animals, easy to measure the yield. Thus, not only those interested primarily in mammary function, but also those interested in the passage of substances from mother to infant and in the excretion of substances from the body in milk, have resorted to the extremely simple experiment of giving a tracer amount (generally an isotope) of a substance to a lactating animal and seeing if it appears in the milk. Unfortunately it is not so easy to extract the maximum amount of information from such a convenient experiment (often, however, employing a very expensive tracer) and those who seek to do so should be reminded that there are traps for the unwary.I cannot mention all the papers (probably hundreds) where isotopes have been used in mammary studies, but I will attempt to refer to work where isotopes have been particularly useful, or have given wrong information through misinterpretation of the data. This should be useful to nutritionists interested in using tracers in their studies. Any organ, including mammary glands, may be regarded as an organism needing to be fed. Thus one may regard what the mammary glands extract from the blood passing through them as their food, which they use partly to nourish themselves and partly to produce milk, and tracers have played an important role in identifying these compounds. The first use of a tracer in lactation was made thirty-five years ago by the pioneer in the use of isotopes in biology, George Hevesy, (Aten & Hevesy, 1938) and, not surprisingly, the paper is a classic. They gave heavy water (D,O) and inorganic 32P to lactating goats and correctly deduced that: ( I ) mammary tissue is more permeable to water than to inorganic phosphate; (2) the source of milk inorganic P, casein P and ester P is the inorganic P of plasma and not plasma phosphatide P, since the specific activities (SA) of milk inorganic P and casein P were similar to the SA of plasma inorganic P and higher than that of plasma phosphatide P. They also buried the old idea that plasma phospholipids are taken up by the udder and hydrolysed, and that the lipid so released is an important precursor of milk fat, because in their experiments tissue phosphatides had a higher SA than plasma phospholipids and therefore must have been formed in the udder and not taken from plasma. Furthermore they made a useful estimate of the rate of milk secrction in terms of the time taken for milk SA to reach the initial plasma SA, which was 3-4 h.Following this excellent start, papers began to appear in which the information obtained from the use of valuable isotope was meagre.