Abstract POULTRY breeders interested in the improvement of egg quality would like to know how many eggs are needed to adequately evaluate the real but unknown inherent breeding worth of a hen. Some breeders feel that observations on two or three eggs per hen are sufficient, while others think that a larger number of eggs should be examined. The key to a problem of this type is based on a knowledge of the repeatability of the quality trait under consideration. Repeatability may be defined for our purposes as the correlation between observations on successive eggs laid by the same hen. The limits of repeatability (R) are between zero and one. If R is large only a few eggs per hen would be required, whereas if R is small several eggs may be required to accurately evaluate the egg quality characteristics of a hen. In a previous paper (Scheinberg et al., 1953) …