Recurrent implantation failure (RIF) is repeated unsuccessful embryo transfers (ETs).

To identify predictive embryonic markers of implantation in RIF, following pre‐implantation genetic screening (PGS) of cleavage stage embryos, after accounting for male and female factors.

Materials and Methods
Retrospective analysis of RIF patients undergoing PGS after correction of modifiable causes.

Eighty‐four patients underwent 140 in vitro ferilisation cycles. Forty‐one cycles were excluded: 12 (no embryo for transfer), four (double ETs) and 25 (no biopsy). Sixty‐three patients underwent 99 single euploid ETs (48 fresh, 51 frozen) resulting in 11 biochemical pregnancies, 36 clinical pregnancies (CP), and six miscarriages and 30 live births (LB). Frozen ET was more successful than fresh; respective live birth rate (LBR) and clinical pregnancy rate (CPR), 39.2% versus 20.8%, (P = 0.02), 45.1% versus 27.1% (P = 0.04).
LBR and CPR were lower when 5–6 blastomeres were present at embryo biopsy, compared to embryos with ≥7 blastomeres: 15.4% versus 32.6% (P = 0.185) and 15.4% versus 39.5% (P = 0.074) respectively. Serum β human chorionic gonadotropin (βhCG) concentration was greater when a more developed embryo was biopsied (r = 0.448, P = 0.017 and r = 0.476, P = 0.118, fresh and frozen transfers, respectively).
Embryo morphokinetic analysis demonstrated faster development to blastocyst stage when more cells were present at biopsy: mean 103.3, 102.2 and 96.0 h for biopsy at the 5–6, 7–8 or ≥9 cell stage respectively (P = 0.040 for difference between 7–8 cells vs ≥9).

After cleavage stage biopsy, frozen ET was more successful than fresh ET. Chance of conception and serum βhCG concentration correlated with number of cells present at time of biopsy.