PURPOSE: To identify the benefit of extending embryo culture until day (D)7 based on patients and cycle characteristics. METHODS: A retrospective cohort study was conducted including 25,120 blastocysts from 5278 PGT-A autologous cycles between 2017 and 2022. A theoretical cumulative live birth rate (CLBR) was calculated by binomial density function. An increase of ≥ 5% in theoretical CLBR was considered a tangible benefit when obtaining ≥ 1 euploid D7 blastocyst and ≤ 3 euploid blastocysts from D5/D6. A predictive model was built considering the number of embryos eligible for extended culture until D7, number of blastocysts already biopsied on D5/D6, and patient's age. RESULTS: Euploidy rates decreased for blastocysts biopsied on D5, D6, and D7 (55.6%, 39.7%, and 27.1%, P <
0.001, respectively). The probability of tangible benefit was increased with more embryos available for extended culture until D7, was decreased with higher D5/D6 blastocysts already biopsied and for older patients. The overall AUC of the final model in the validation sets was 0.75 (95% CI 0.72-0.78). Based on the predictive model, in poor cycles (<
1% tangible benefit), the benefit rate from extended culture was 0.3% and for moderate, good, and best cycles (1-10%, 10-20%, and ≥ 20% tangible benefit) were 4.4%, 14.0%, and 29.3%, respectively. An application of the predictive model is available online for external testing: https://artfertilityclinics.shinyapps.io/WET-D7/ . CONCLUSION: The predictive model provides a decision-making tool to objectively identify cycles that would benefit from extending embryo culture until D7.