Embryos were recovered in vivo from donor ewes at day 4 and transferred into superovulated unmated recipient ewes given an injection of PMSG (1600 IU) at day 13.5 of the preceding cycle. The recipient ewes were slaughtered at either 5 (group 1) or 8 (group 2) days after transfer. The recovered blastocysts were transferred back into the original donor ewes and pregnancy was allowed to continue until term. In order to observe the effect of the two transfers on blastocyst viability, the recipient ewes were not superovulated in group 3. Only one transfer was carried out at day 4 in group 4, and then pregnancy was allowed to continue in the superovulated recipient ewes. From day 3 to day 8, 12 or 20 (groups 1, 2 + 3 and 4, respectively), the peripheral blood of recipient ewes was sampled once a day for progesterone assay and four times a day for estradiol-17β assay. At 9 or 12 days, 50, 62 and 68% of the transferred embryos were recovered in groups 1, 2 and 3, respectively. These rates were not statistically different from the pregnancy rate in group 4 (64%). After the second transfer, 43, 54 and 40% of the blastocysts developed into lambs (groups 1, 2 and 3, respectively). There was no statistical difference between these results. However, as we noted in previous studies, in spite of the changes in the uterine medium caused by superovulation and which accelerated blastocyst development, the uterus of superovulated ewes could assume pregnancy. The first transfer decreased the number of pregnant ewes to 65% and the second transfer lowered the number of blastocysts giving lambs to 50%. The level of progesterone varied considerably in recipient ewes giving lambs. When the level of progesterone was low at D4, one embryonic mortality was recorded. The level of estradiol-17β showed large variations and seemed to have no relation to blastocyst survival.