What factors most strongly predict IVF success?

[Partially verified against ESHRE Ovarian Stimulation Guideline 2026] ESHRE (2026) draws a distinction that is directly relevant here and frequently collapsed in practice. For predicting ovarian RESPONSE: either AFC or AMH is recommended (strong). Age, BMI, basal FSH, inhibin B, basal oestradiol, basal progesterone and basal LH are not recommended for this purpose (strong). For predicting PREGNANCY and LIVE BIRTH: female age and BMI are predictors (strong). AFC, AMH, basal FSH, basal LH, basal oestradiol, basal progesterone and inhibin B are NOT recommended for predicting pregnancy and live birth (strong). So AMH predicts how many oocytes you will get. It does not predict whether she will have a baby. That distinction should govern how ovarian reserve results are communicated — a low AMH is a stimulation planning parameter, not a prognosis, and it is routinely delivered as the latter. Female age remains the dominant predictor of live birth, principally through embryo euploidy. Other factors with recognised influence, not individually graded in that guideline: number of oocytes retrieved and embryos available, embryo quality and ploidy status, uterine cavity status, previous treatment response, BMI, smoking, and duration and cause of infertility. Reviewer: if quantitative prediction is offered to patients, it should come from a validated prediction model with its population stated, or from your own outcome data — not from a general list of factors.