[How can we nowadays select the best embryo to transfer?].

Multiple pregnancies stand as the most common adverse outcome of assisted reproduction technologies (ART) and the dangers associated with those pregnancies have been reduced by doing elective single embryo transfers (e-SET). Many studies have shown that e-SET is compatible with a continuously high pregnancy rate per embryo transfer. Yet, it still becomes necessary to improve the selection process in order to define the quality of individual embryos - so that the ones we choose for transfer are more likely to implant. First, analysis of embryo morphology has greatly helped in this identification and remains the most relevant criterion for choosing the embryo. The introduction of time-lapse imaging provides new criteria predictive of implantation potential, but the real contribution of this system - including the benefit/cost ratio - seems to be not yet properly established. In this context, extended culture until blastocyst stage is an essential practice but it appears wise to keep it for a population showing a good prognosis. Then, the failure of aneuploid embryos to implant properly led to achieve preimplantation genetic screening (PGS) in order to increase pregnancy and delivery rates after ART. However, PGS by fluorescence in situ hybridization (FISH) at day 3 is a useless process - and may even be harmful. Another solution involves using comparative genomic hybridisation (CGH) and moving to blastocyst biopsy. Finally, it is envisaged that morphology will also be significantly aided by non-invasive analysis of biomarkers in the culture media that give a better reflection of whole-embryo physiology and function.