The amount and distribution of prenatal mortality was estimated in 3 groups of does; Group I, a series of 126 1st and 2d pregnancies in 73 control does; Group 2, 15 superovulated and 16 control does, and Group 3, 16 does having 5 ova transferred to 1 uterine horn and either 15 or 20 ova transferred to the other uterine horn. Approximately 5% of the untreated does suffered total loss of their ova before implantation. In does in which implantation occurred, the loss amounted to 9.7% of the ova before implantation and 18% of the embryos after implantation. After implantation the loss was distributed as follows: immediately after implantation, 7.0%; days 8 to 17, 65.8%, and days 17 to 23, 27.2%. In superovulated does, having a mean of 28.7 implantations, no reliable estimate can be given for the pre-implantation loss; after implantation 79% of the embryos were lost compared with 22.6% in the controls and the mortality tended to occur earlier in the superovulated does. In the Group 3 recipients, there was no significant difference between the 2 uterine horns in the loss of ova before implantation. After implantation, 23% of the embryos were lost on the side with few implantations compared with 64% on the crowded side. It was concluded that "local" factors (deficiency in placental development or inadequacy of vascular supply) were responsible for the abnormally heavy post-implantation loss in uteri with an excessive number of implantations. It was suggested that there exists a "ceiling value" for the number of implantation sites that can be maintained successfully.