Abstract Study question Does application of a Freeze All Embryo (FAE) policy achieve superior live birth rates in practice compared to fresh IVF transfer cycles? Summary answer FAE FETs achieved significantly higher LBRs than fresh for both the first ET per cycle and cumulatively per cycle. What is known already The efficacy of vitrification methods has led to segmentation between the ovarian stimulation and embryo transfer cycle becoming increasingly applied. Randomised controlled trials have shown that for high responders, especially those at risk of ovarian hyperstimulation syndrome, the FAE approach is likely to be favourable. However, whether a routine freeze all approach applied in routine clinical practice improves outcomes in a broader group of patients remains unclear. Study design, size, duration This is a retrospective analysis of 7,647 autologous fresh oocyte retrievals and 9,609 subsequent ETs, performed between 2016-2022. The primary study aim was to compare (cumulative) live birth rates between fresh and frozen thawed embryo transfers after FAE IVF cycles. The groups consisted of 3,707 fresh IVF transfers plus 1,533 subsequent FETs derived from those cycles, versus 3940 FAE cycles with 3969 subsequent FETs. Participants/materials, setting, methods 5,808 patients underwent fresh oocyte retrievals (2016-2022) at a single UK-based centre. In total, 4,997 of these patients underwent 9,609 ETs; either fresh or frozen. For analysis, FAE cycles and fresh IVF cycles underwent 1:1 full propensity score matching for age and number of eggs collected, using probit regression of treatments on covariates. Logistic regression analyses were fit including full matching weights in the estimation to compare LBR superiority between groups. Main results and the role of chance Patients undergoing fresh transfers were significantly older and yielded significantly fewer oocytes per retrieval (37 and 36, P < 0.001; 7 ± SE 0.07 and 12 ± SE 0.1, P < 0.001, respectively). After full 1:1 matching for age and oocytes collected, the means for both groups were 36 and 15, respectively. Standardised mean differences after matching was <0.2. The first FET from FAE cycles achieved significantly superior LBRs and was a significant predictor of LB compared to fresh ETs when adjusted for covariates (37% and 26%, respectively, aOR 1.3; 95% CI 1.2-1.5, P < 0.001). Cumulative LBs achieved per cycle were also significantly higher for FAE cycles (56% and 37%, respectively). When stratifying by age, first ET LBRs between groups were comparable for <35s (FAE 42%, fresh 40%). However, with increasing age, FAE first FETs achieved higher live births than fresh transfers (35-37, 36% and 33%; 38-39, 34% and 23%; 40-42, 27% and 14%; 43-45, 15% and 5%, respectively). Limitations, reasons for caution This is a retrospective analysis and, although cohorts underwent full 1:1 matching, inherent bias may be present within the data and must be considered when drawing conclusions. Wider implications of the findings These findings suggest that in clinical practice, FETs after a FAE cycle may achieve superior LBRs compared to fresh ETs, particularly in older patients and for higher responders. Overall, FAE cycles were observed to result in higher cumulative LBRs per cycle than cycles that begin with a fresh transfer. Trial registration number Not applicable