Egg-weight apportionment patterns and chick survival were investigated in two-egg clutches of Caspian terns(Sterna caspia) in the Great Lakes. First-laid eggs (A eggs) were typically heavier than second-laid eggs (B eggs) and hatched an average of 1.8 days earlier. The A-egg fraction of total clutch weight increased with total clutch weight. Increased egg weight did not increase the probability of hatching, but did increase the fledging success of A chicks: specifically, A eggs that produced fledglings were significantly heavier than those whose chicks did not fledge. No such relationship held for B eggs. The main effect of B-egg weight was in prolonging life: excluding those few that fledged, egg weight was correlated significantly with age at mortality. Chick mortality appeared to be due mainly to starvation and gull predation. We suspect that greater body size and (or) behaviour associated with body size helps chicks escape risk from gape-limited predators. The hatching of A eggs had a negative effect on the survival of B chicks, but the reverse was not true. B offspring may act as insurance against the loss of A siblings. The potential benefit to parents of investing more in B eggs appears to be constrained by sibling competition with A chicks, which often results in brood reduction. The pattern of egg-weight apportionment in eggs is interpreted as a parental response to the differences in the reproductive values of asynchronously hatching chicks.