In six species of dimorphic raptors (females larger than males) and one passerine (males larger than females), the sex ratio at fledging varied systematically with brood size at fledging. In all species the strongest bias toward the smaller sex was established in the largest as well as the smallest broods; a more even distribution of males and females was observed in broods of intermediate size. We explored a specific differential mortality explanation for this sex ratio variation. Our hypothesis postulates that variation in mortality is caused by differences in food demand between broods of the same size, due to their sex composition. Data from the marsh harrier Circus aeruginosus on gender-related food demand and overall nestling mortality were used to predict the frequency of surviving males and females at fledging, assuming an even sex ratio at hatching and random mortality with respect to both sexes within broods. The model quantitatively fits the marsh harrier data well, especially in broods originating from large clutches. Although we anticipate that other mechanisms are also involved, the results support the hypothesis of sex-ratio-dependent mortality, differential between broods, as the process generating the observed broad-size dependence of fledgling sex ratios in sexually dimorphic birds.