Abstract Study question Does longitudinal growth of euploid embryos differ from aneuploids? Summary answer Euploid embryos had significantly lower growth variance score (GVS), indicating steadier growth when observed with time-lapse monitoring, compared to aneuploid embryos. What is known already A major quest for optimizing embryo selection is the identification of morphokinetic parameters to distinguish euploid from aneuploid embryos. Although some studies have identified altered morphokinetic parameters related to embryo ploidy, outcomes were not reproducible, and could not provide sufficient evidence to apply it. Moreover, longitudinal changes during this temporal process of embryo development, their reference standards, and the impact of such changes on embryonic fate remains unknown. Aiming to develop a growth variance model, using the reference ranges of embryonic morphokinetic growth as a single growth trajectory might provide a better additive value to distinguish euploid from aneuploid embryos. Study design, size, duration This observational, retrospective cohort study was performed in a tertiary referral IVF centre between December 2020 and March 2023. The study included a total of 677 blastocysts from 147 cycles tested for Preimplantation Genetic Testing for aneuploidy (PGT-A) by next generation sequency (NGS). All embryos were monitored by time-lapse (TL) system during culture. Participants/materials, setting, methods Growth variance score (GVS) is a measure of embryonic growth pattern calculated using the TL morphokinetic reference ranges (from tPB2 until tEB), developed for euploid embryos. Higher GVS scores indicate more erratic growth while lower scores indicate steadier growth. Trophectoderm (TE), inner cell mass (ICM), and expansion grade were assessed immediately before biopsy using Gardner’s criteria. Euploid and aneuploid blastocysts were compared in terms of ICM, TE grade, day of biopsy and GVS score. Main results and the role of chance Among tested blastocysts, 383 were euploid (56.6%) and 294 of them were abnormal (43.4%). Euploid embryos had significantly lower GVS compared to aneuploid embryos (mean difference: 0.30 lower, 95% CI: 0.16 to 0.45; P < 0.001). Among day 5 biopsies, euploids had significantly lower GVS compared to aneuploids (mean difference: 0.27 lower, 95% CI: 0.04-0.51; P = 0.019). Among Day 6+ biopsies, euploids also had lower GVS compared to aneuploids, but the difference did not reach statistical significance (mean difference: 0.09 lower, 95% CI: -0.08-0.28, P = 0.30). Regression analysis showed GVS was significantly and independently associated with euploidy after adjusting for day of biopsy and morphological grade (aOR: 0.83, 95% CI: 0.70 – 0.99; P = 0.048). We compared two embryo selection modalities, the first one using morphology and biopsy day, the second one incorporating GVS to morphology and biopsy day. In 90% (95% CI: 87.2 – 92.8%) of the cycles, both modalities made a concordant selection by choosing the same embryo as the best. In the remaining 10% (95% CI: 6.4% - 14.7%) of the cycles, they made a discordant choice and the euploidy rate of embryos chosen by the combination of GVS and morphology were higher by 20.4% (95% CI: -0.06 to 44.0%, P = 0.060). Limitations, reasons for caution To confirm its reproducibility in other clinical and laboratory settings, GVS should be replicated with larger sample sizes and in external cohorts. Wider implications of the findings The GVS model might emerge as a valuable prognostic tool to select euploid blastocysts within patient’s own pool of embryos. GVS could be easily incorporated into laboratory work-flow when a time-lapse monitor is available to prioritize embryos for transfer, given external validations are successful. Trial registration number Not applicable