It is widely held that the first two blastomeres of mammalian embryos are equally totipotent and that this totipotency belongs to the group of regulative properties. However, this interpretation neglects an important aspect: evidence only came from successful monozygotic twins which can speak only for those pairs of half-embryos that are able to regulate in the first place. Are the frequently occurring incomplete pairs simply an artefact, or do they represent a real difference, be it in the imperfect blastomere's ability to regulate growth or in the distribution of any compound X that constrains regulation? Using the model system of mouse embryos bisected at the 2-cell stage after fertilization, we present evidence that the interblastomere differences evade regulation by external factors and are already latent in oocytes. Specifically, an interblastomere imbalance of epiblast production persists under the most diverse culture conditions and applies to the same extent in parthenogenetic counterparts. As a result, cases in which twin blastocysts continued to develop in only one member account for 65 and 57% of zygotic and parthenogenetic pairs, respectively. The interblastomere imbalance is related to the subcellular distribution of gene products, as documented for the epiblast-related gene Cops3, using mRNA FISH in super-resolution mode confocal microscopy. Blastomere patterns of Cops3 mRNA distribution are α-amanitin-resistant. Thus, the imbalance originates not from de novo transcription, but from influences which are effective before fertilisation. These data expose previously unrecognized limits of regulative capacities of 2-cell stage blastomeres and point to aspects of cytoplasmic organization of the mouse oocyte that segregate unequally to blastomeres during cleavage.