Human embryo culture media, intended for assisted reproductive technologies (ARTs), are released for clinical use if they pass the mouse embryo assay (MEA). This assay prescribes that at least 70% of in vivo fertilized mouse 1-cell embryos form blastocysts, in order to grant the culture medium approval. In the fertility clinic, however, human embryos undergo more manipulation than their MEA counterparts through, for example, fertilization by intracytoplasmic sperm injection (ICSI); further, only a minority of the embryos transferred to the uterus goes on to establish gestations. In this context, we asked if the results of the MEA only depend on the type of in vitro culture, or are also affected by the method of fertilization. Superovulated B6C3F1 mouse oocytes were fertilized by ICSI using C57Bl/6 sperm. Pronuclear-stage eggs were allocated to four developmental environments: two ART culture protocols (HTF/MultiBlast, Irvine Scientific; ISM1/ISM2, Origio), standard mouse culture medium (KSOM(aa), made in-house) and the oviduct of pseudopregnant CD1 mice. As control for the invasive manipulation, pronuclear-stage eggs were generated by mating (B6C3F1 × C57Bl/6) and cultured in KSOM(aa) medium. Embryos were recovered from culture or from the CD1 uterus and scored for blastocyst formation at 96 h of development (Table 1). For these blastocysts, we determined the number of total, inner cell mass (ICM), and trophectoderm (TE) cells (Table 1) by confocal immunofluorescence microscopy (Schwarzer et al. 2012 doi:10.1093/humrep/des223). Our results show that ART culture protocols applied to mouse ICSI embryos are not equivalent in supporting blastocyst formation. Based on blastocyst rates, the ranking observed here after ICSI, reflects the ranking reported by us for IVF embryos (Schwarzer et al. 2012); that is, KSOM(aa) > HTF/MultiBlast > oviduct > ISM1/2. This similarity suggests that the effect of in vitro culture on mouse development exceeds the effect of ICSI, provided gametes are of good quality. From the analysis of cell numbers, we note that while the ICM/TE ratios are not of easy interpretation, the absolute numbers of cells in the ICM draw a clear line between the environment of the oviduct and those of culture media. Irrespective of the ICM/TE ratio, only the oviduct environment secures 8 cells in the ICM (Table 1). Soriano and Jaenisch (1986 Cell 46, 19–29) reported that 8 cells of the ICM are set aside to give rise to the body of a mouse. In summary, the current MEA is a valuable assay to assess the quality of culture medium, however, its refinement is necessary to better model the adaptive properties of embryo culture when different methods of fertilization are applied. Until the MEA is extended into postimplantation development, as we advocate (Schwarzer et al. 2012), the absolute numbers of cells in the ICM may be a better gauge of embryo quality than the blastocyst rates. Table 1.Mouse embryo assay outcomes after ICSI