Abstract In rodents, steroids are able to pass between fetuses. Although not previously examined in mice, we have provided evidence that blood flow in the uterine loop artery in rats is bidirectional and that steroids are transported between fetuses by diffusion via the uterine lumen, not via the uterine blood vessels of the mother. The direction of blood flow in the uterine loop artery and vein feeding each uterine horn in house mice was examined on day 17 of pregnancy. Dye was injected into the heart to determine the direction of blood flow in the uterine artery while injection of dye into individual placentae was used to determine the direction of blood flow in the uterine vein. Blood entered the loop artery from both the dorsal and caudal ends and was thus bidirectional. Venus blood flow from placentae was in a rostral direction from placentae in the rostral portion of a uterine horn and in a caudal direction from placentae in the caudal portion of a uterine horn. Comparison of anogenital distance and body weight at birth using a variety of classification schemes, based on different assumptions about the mechanism and direction of transport of steroids between fetuses, showed that the only scheme which accounted for variation in anogenital distance at birth in female mice was that which was based on the number of directly adjacent male siblings in utero. Taken together with recent findings in rats reported elsewhere, we conclude that blood flow in the uterine blood vessels is bidirectional, and steroids are transported between fetuses by diffusing through the amniotic fluid and across the fetal membranes of adjacent fetuses. We found that the common method of using the ratio of anogenital distance/body weight can lead to erroneous conclusions concerning intrauterine position effects on anogenital distance. Analysis of covariance tests whether variation in body weight contributes to differences in anogenital distance due to intrauterine position, and this method of analysis should be used in future studies.