Previous attempts to recapitulate embryogenesis in a developmentally relevant context started with aggregates composed of a few thousand ESCs, termed embryoid bodies (EB), that upon induction of differentiation reveal a surprising level of autonomous cell fate patterning, in some cases giving rise to spatially arranged ectoderm, mesoderm and endoderm layers. However, cell fate patterning and morphogenesis in EBs is poorly reproducible and cannot be controlled. These limitations necessitate further research on the self-organization potential of ESCs, and in particular the development of more robust in vitro models mimicking post-implantation mouse embryos. This PhD thesis explored three novel approaches towards this goal. Firstly, together with collaborators, I focused on a previously established in vitro model of gastrulation termed gastruloids. Gastruloids emerge from small ESC aggregates that, upon activation of Wnt signaling, break symmetry and undergo axial morphogenesis to create structures that are similar to the developing tail of the mouse embryo. A systematic improvement of the existing culture conditions allowed gastruloids to be cultured for an extended period of time, resulting in an embryo-like patterning along the three major body axes. Furthermore, gene expression analyses, combined with immunohistochemistry and in situ hybridization assays, demonstrated that gastruloids exhibit spatio-temporal activation of Hox genes in a way that is remarkably similar to post-implantation mouse embryos. Secondly, I report a novel gastruloid culture approach, which for the first time promotes the formation of anterior neural tissues. Mouse ESCs were aggregated and treated in high-throughput in a bioengineered microwell array system to generate homogeneous ‘artificial epiblasts’ that transcriptionally and morphologically resemble post-implantation epiblasts. These epiblast-like (EPI) aggregates break symmetry and undergo axial elongation to establish antero-posterior patterning. I provide evidence for attenuated Wnt signaling and the presence of an epithelium in the starting EPI aggregates as the major determinants for the formation of anterior neural tissues, that are completely absent in conventional gastruloids. Lastly, I systematically explored the role of extraembryonic tissues in the patterning of artificial epiblasts. By combining EPI aggregates with trophoblast stem cell aggregates, hybrid self-organizing structures termed EpiTS embryoids were generated that mimic epiblast and extraembryonic ectoderm, respectively. EpiTS embryoids were found to execute an elaborate self-organization program that is initiated through highly localized gastrulation-like processes at the embryonic-extraembryonic interface. Upon axial elongation, EpiTS embryoids give rise to primitive brain-like and preplacodal ectoderm-like regions. These observations highlight the role of extraembryonic tissues for proper induction of gastrulation and in particular the formation of anterior neural tissues. Altogether, this thesis introduces three innovative and highly versatile stem cell culture technologies to study post-implantation mouse embryonic development. The approaches are amenable to large-scale studies aimed at identifying novel regulators of gastrulation and anterior neural development that is currently out of reach with existing experimental tools. This work should contribute to the advancement of the nascent field of synthetic embryology.