Genomic imprinting of the largest known cluster, the Kcnq1/KCNQ1 domain on mChr7/hChr11, displays significant differences between mouse and man. Of the fourteen transcripts in this cluster, imprinting of six is ubiquitous in mice and humans, however, imprinted expression of the other eight transcripts is only found in the mouse placenta. The human orthologues of the latter eight transcripts are biallelically expressed, at least from the first trimester onwards. However, as early development is less divergent between species, placental specific imprinting may be present in very early gestation in both mice and humans.Human embryonic stem (hES) cells can be differentiated to embryoid bodies and then to trophoblast stem (EB-TS) cells. Using EB-TS cells as a model of post-implantation invading cytotrophoblast, we analysed allelic expression of two telomeric transcripts whose imprinting is placental specific in the mouse, as well as the ncRNA KCNQ1OT1, whose imprinted expression is ubiquitous in early human and mouse development. KCNQ1OT1 expression was monoallelic in all samples but OSBPL5 and NAP1L4 expression was biallelic in EB-TS cells, as well as undifferentiated hES cells and first trimester human fetal placenta. DCN on hChr12, another gene imprinted in the mouse placenta only, was also biallelically expressed in EB-TS cells. The germline maternal methylation imprint at the KvDMR was maintained in both undifferentiated hES cells and EB-TS cells.The question of placental specific imprinting in the human has not been answered fully. Using a model of human trophoblast very early in gestation we show a lack of imprinting of two telomeric genes in the KCNQ1 region and of DCN, whose imprinted expression is placental specific in mice, providing further evidence to suggest that humans do not exhibit placental specific imprinting. The maintenance of both differential methylation of the KvDMR and monoallelic expression of KCNQ1OT1 indicates that the region is appropriately regulated epigenetically in vitro. Human gestational load is less than in the mouse, resulting in reduced need for maternal resource competition, and therefore maybe also a lack of placental specific imprinting. If genomic imprinting exists to control fetal acquisition of maternal resources driven by the placenta, placenta-specific imprinting may be less important in the human than the mouse.