Abstract Study question Does a physiological gradient of oxygen concentration in embryo culture atmosphere yield a higher percentage of fertilized oocytes developing into morphologically optimal blastocysts? Summary answer Study favours sustaining a static 5% oxygen level, instead of a gradient, during extended embryo culture, resulting in more viable embryos for cryopreservation and transfer. What is known already To date, there’s ongoing research about the optimal oxygen concentration for embryo culture. The available data from studies based on animal models show that the O2 level in the fallopian tube is around 8%, and the in-utero O2 level is at around 2% (Wale & Gardner, 2010). Data on oxygen measurement in the human uterus is scarce, yet aligns with similar findings in other mammalian species (Ng et al., 2018). These findings support the use of sequential O2 levels for embryo culture. Our study stands as the first to implement such dynamic oxygen culture conditions within an in vitro setting. Study design, size, duration A prospective sibling-split oocyte study, analyzing 44 ICSI cycles from January 2022 to January 2023, with alternate allocation into a static 5% (Control group) during entire culture period or under an oxygen gradient (intervention group), starting with 8% from day-0 to day-3, continuing with 5% on day-3 and following with 2% of oxygen from the end of day-3 to day-5/6. Participants/materials, setting, methods 44 ICSI cycles, which produced a total of 521 metaphase II oocytes. A total of 353 2PN embryos were cultured and annotated in time-lapse for morphokinetic and morphometric analysis. Main results and the role of chance While morphological outcomes slightly favored the intervention group for day 2 and day 3 embryos, the trend reversed by day 5. The control group (5% O2) showed a notably higher proportion of zygotes developed to clinically utilized blastocysts (47% vs. 37%, p = 0.016) and consequently more of them were used for transfer on day 5 (10% vs. 4%, P = 0.008) compared to the intervention group. Additionally, though not statistically significant, the control group showed a higher percentage of morphologically optimal blastocysts on day 5 (10.3% vs. 14.7%, p = 0.1233). Morphokinetic timings, as revealed by time-lapse analysis, indicated a delayed development in the intervention group, notably seen in the initiation of blastulation (p = 0.0024), initiation of expansion (p < 0.0001), and full blastocyst formation (p < 0.0001). The speed of blastulation, however, does not necessarily reflect differences in implantation potential, as implantation (58.6% vs 57.1% p > 0.92) and clinical pregnancy rates (55.2% vs 57.1%, p > 0.92) after transferring blastocysts from either group were comparable. Limitations, reasons for caution Measuring the in vivo O2 levels in female reproductive tract should take into account possible oscillations and dynamic nature of gas and nutrient exchange. It’s also important to acknowledge the complexity of embryo’s adaptive response, making it challenging to attribute observed outcomes solely to the impact of targeted oxygen concentration. Wider implications of the findings Based on these clinical results, we will analyze the metabolomic footprint in spent embryo culture under static 5% or gradient 8-5-2% oxygen concentration, employing non-invasive NMR and chromatography methods. A randomized control trial should be conducted to explore the effect of different oxygen regimes on clinical outcomes. Trial registration number NCT05898178