Measures of repeatability have long been used to assess patterns of variation in egg size within and among females. We compared different analytical approaches for estimating repeatability of egg size of Black Brant. Separate estimates of repeatability for eggs of each clutch size and laying sequence number varied from 0.49 to 0.64. We suggest that using the averaging egg size within clutches results in underestimation of variation within females and thereby overestimates repeatability. We recommend a nested design that partitions egg-size variation within clutches, among clutches within females, and among females. We demonstrate little variation in estimates of repeatability resulting from a nested model controlling for egg laying sequence and a nested model in which we assumed laying sequence was unknown.