1. An analysis has been made of the association between transferrin genotype and progeny test for both milk yield and fat content in a sample of 879 bulls of five dairy cattle breeds, the majority used in artificial insemination in Great Britain. Although significant effects were found in only one of the ten analyses when the breeds were considered separately, a significant effect on milk yield was found when the results from all breeds were combined. A significant effect was not found for fat content, although the effects of the allele substitutions were in the opposite direction to that on milk yield in five cases out of six.

2. Although our results were in the same direction as those of Ashton, the effects were smaller. It was estimated that the locus accounted for 1·1% of the genetic variation in yield and 0·4% of the genetic variation in fat content.

3. Some information based on the production records of 178 experimental cows with known genotypes was also analysed. For milk yield, the pattern of effects was similar to that in the sire data though the differences were much larger. The variance removed by fitting constants was not statistically significant for milk yield or fat content. It was estimated that the sire data contained some fifty times as much information as did the cow data.