Abstract Study question How does ICSI-induced leakage of trophectoderm (TE) or inner cell mass (ICM) from microslits in the zona pellucida (ZP) affect pregnancy outcomes following blastocyst transfer? Summary answer Microcellular leakage of TE and ICM did not affect pregnancy outcome, but blastocysts with ICM leakage above 53.6% had a lower pregnancy rate. What is known already The leakage of TE cells through the ZP (trophectoderm vesicles; TV) has been reported to be more common in ICSI than in conventional IVF, caused by the opening of the ZP with microslits for ICSI. This phenomenon may be improved through the implementation of assisted hatching during transplantation, however, there are few reports investigating its impact on pregnancy.Furthermore, there are no reports on the impact of inner cell mass leakage (inner cell mass vesicles; IV) on pregnancy outcomes, and it is unclear what level of ICM leakage is acceptable for transplantation. Study design, size, duration In this retrospective study, we analyzed clinical medical reports at the Takahashi Women’s Clinic in Japan. A total of 551 patients (1093 blastocysts) who underwent freeze-thaw single blastocyst transfer with laser assisted hatching between January 2019 and October 2023 were included. Blastocysts used for transfer were 2PN embryos obtained by PIEZO-ICSI and cultured in a time-lapse incubator (EmbryoScope+). Participants/materials, setting, methods The blastocysts were classified into control (N = 682), TV (N = 377), and IV groups (N = 34). Pregnancy outcomes were compared using logistic regression analysis, considering patient age, blastocyst grade (Gardner criteria), culture days, BMI, and basal AMH. Bonferroni correction was used for multiplicity correction, considering P < 0.0167 as significant. The ICM leakage rate (%) was calculated as: (leaked ICM area/ICM area before leakage) × 100. Pregnancy cutoff was determined by ROC analysis. Main results and the role of chance The hatching rates for the control, TV, and IV groups were 69.6% (475/682), 78.8% (297/377), and 82.4% (28/34), respectively. Pregnancy rates (fetal heartbeat) were 44.6% (304/682), 57.0% (215/377), and 50.0% (17/34), respectively, and miscarriage rates were 26.3% (80/304), 26.0% (56/215), and 23.5% (4/17), respectively. Logistic regression analysis considering patient background showed no significant differences in hatching, pregnancy, and miscarriage rates (P > 0.05). Furthermore, mean weight, sex ratio, and congenital anomaly rates for the 107 and 7 children born in the TV and IV groups, respectively, were not significantly different than those for the 175 children born in the control group. In examining ICM leakage, the mean area before leakage was 3162.6 µm2 and mean area of leaked ICM was 1292.1 µm2 in the IV group, with an average of 40.9% of ICM leaking outside the clear zone. ROC analysis showed that the cutoff for ICM leakage rate for pregnancy was 53.6%, with significantly lower pregnancy rates when the ICM leakage rate was ≥53.6% (18.2% [2/11] vs 65.2% [15/23], p = 0.0255, Fisher’s exact test). Limitations, reasons for caution This study was conducted at a single in vitro fertilization center. Pregnancy, miscarriage, and fetal anomaly rates at birth related to embryo transplantation are based on ongoing pregnancy dates, and complete live birth data for all pregnancies are not yet available. Wider implications of the findings No differences in pregnancy outcomes or fetal anomalies were found among groups. However, pregnancy rates decreased when ICM leakage rates were over 53.6%. Blastocysts with IV may successfully establish a pregnancy, but considering the leaked ICM proportion is crucial. If considered for transplantation, obtaining sufficient informed consent is essential. Trial registration number not applicable