X-chromosome inactivation (XCI) in female mammals enables dosage compensation for X- linked gene products between the sexes. Since its discovery fifty years ago (Lyon, 1961), the developmental regulation of this process has been extensively investigated in mice (see Morey and Avner,2011; Augui et al, 2011; Gendrel and Heard, 2011 for reviews), but hardly at all in non-murine species. In mice, the X chromosome of paternal origin (Xp) is silenced during early embryogenesis due to imprinted expression of the regulatory RNA, Xist (X- inactive-specific-transcript). Paternal XCI initiates early in mice and is then reversed in the inner cell mass (ICM) of the blastocyst. Random inactivation of either paternal or maternal X then ensues in epiblast cells. We investigated the developmental regulation of XCI in rabbit and human embryos (Okamoto et al, 2011) and were able to show that in these mammals, Xist is not subject to imprinting and X inactivation begins much later than in the mouse. Furthermore, Xist is up-regulated on both X chromosomes in a high proportion of rabbit and human embryo cells, even in the ICM. In rabbits, we showed that this triggers XCI on both X chromosomes, implying that the choice of which X chromosome will finally become inactive occurs downstream of Xist up-regulation, a situation that is totally different to the mouse, where choice occurs upstream of Xist RNA accumulation. In humans, on the other hand, XCI is not triggered, even by the blastocyst stage, despite the up-regulation of Xist. Xist is thus expressed, rather than repressed, in Oct4/Nanog positive ICM cells, in both non- murine species. In fact, the status of the two X chromosomes in the ICM is very different in rabbit, human and mouse embryos. In human ICM cells, the two X chromosomes seem to be active, while in the rabbit ICM, the two Xs are initially active (day 4) but then XCI initiates one day later (day 5) in the ICM. This is exactly the opposite situation to that found in the ICM of mouse blastocysts, where the paternal X is initially inactive and then becomes reactivated in the epiblast cells of the ICM. All these results demonstrate the remarkable diversity of X-inactivation initiation mechanisms and highlight major differences between mammals in the requirement for dosage compensation during early embryogenesis.