In rats, dams that conceive in their postpartum oestrus and then lose their firstborn litter bias the sex ratio of the litter toward females in utero. The present study identifies the source of litter sex ratio bias in these postpartum pregnant non-lactating dams. The female bias arises first through the postconception loss of embryos, and second, the loss occurs in midpregnancy between the attachment of the blastocyst to the uterine wall on day 5 and full metrial gland development on day 14. Some pregnancies were restricted to one uterine horn to see if this loss (and thus the opportunity for litter sex ratio biasing) was influenced by local factors operating within the uterine horn. Embryonic loss was more closely associated with the number of embryos implanting in a single horn than with the number implanting in the litter, demonstrating that local crowding within a horn is sufficient for the preferential loss of male embryos. This loss did not cause an obvious decrease in the size of the live-born litter because only those horns with a surfeit of embryos lost them. This process was the same in the right and left horns; both carried and lost the same numbers of embryos. A dam that conceives in her postpartum oestrus and then loses her suckling litter forgoes the implantation delay and uterine healing caused by lactation. Male embryos are less successful at implanting in a uterus only recently vacated by a previous litter.