How is recurrent implantation failure currently defined?

There is no universally accepted definition, and this is a substantive problem rather than a semantic one. Historically RIF has been defined by a threshold — commonly three or more failed transfers of good quality embryos. The difficulty is that an arbitrary count treats very different situations as equivalent: three failed transfers of untested embryos in a 41-year-old and three failed euploid transfers in a 31-year-old are not the same clinical entity, and pooling them has made the literature hard to interpret. The direction of expert opinion has moved toward a probability-based framing: RIF is better understood as failure to implant despite a cumulative expected implantation probability high enough that failure is genuinely unexpected. That depends on embryo number, embryo quality or ploidy status, and maternal age rather than on a transfer count alone. Practical implication for counselling: before investigating RIF, calculate what the expected cumulative probability actually was. A substantial proportion of patients labelled with RIF have simply not yet had enough good-quality embryos transferred, and the appropriate response is continuing treatment rather than investigating. Reviewer: please confirm the current ESHRE working group position and whether a specific definition should be quoted here.