Unilateral ovariectomy in the mouse produces little if any reduction in the total number of eggs shed. Almost all the resulting embryos implant in the uterine horn of the intact side, which is therefore almost twice as crowded as normal. However, the operation does not alter the very significant inverse regression of gestation length on litter size. The duration of gestation in the mouse must therefore be influenced by the total number of conceptuses in the female, rather than by the number of conceptuses in the more crowded uterine horn.Occluding one Fallopian tube, which reduces the number of foetuses to about half but does not affect the number of corpora lutea, is also without effect on the inverse regression of gestation length on litter size. The duration of gestation in the mouse must therefore be related to the number of conceptuses directly, rather than through the number of corpora lutea.The results are consistent with the view that some humoral factor from the placentae and/or foetuses regulates the duration of pregnancy. An analysis of the literature on the endocrinology of pregnancy in the mouse strongly implicates oestrogens. This suggestion must remain speculative until the presence and nature of placental oestrogens in the mouse have been studied.The inverse relation between duration of gestation and litter size is well documented in many species. The sow, however, is a notable exception, since not one of several studies has demonstrated the phenomenon. The reasons for this are still obscure.