OBJECTIVE: We sought to determine whether the early luteal serum progesterone (P4) level predicts the success of IVF treatment with oral dydrogesterone for luteal support. METHOD: This retrospective monocentric cohort study included 242 women who underwent IVF treatment with fresh embryo transfer (ET) between July 2017 and June 2018. The population was unselected, and women were treated according to our unit's usual stimulation protocols. For the luteal phase support (LPS), all women were supplemented with a 10 mg three-times-daily dose of oral dydrogesterone beginning on the day of oocyte pick-up (OPU). Blood sampling was performed on the day of ET (Day 2-3 after OPU) to determine the early luteal serum progesterone level. RESULTS: ROC curve analysis allowed us to determine two thresholds for the prediction of live birth using the early P4 level. Women who had early luteal P4 levels greater than 252 nmol/l had a significantly higher live birth rate (27.1%) than women with early luteal P4 between 115 and 252 nmol/l (17.2%) and women with early luteal P4 below 115 nmol/l (6.0%; p = 0.011). After a multiple regression analysis, an early luteal P4 level greater than 252 nmol/l was still associated with a higher chance of a live birth than a P4 between 115 and 252 nmol/l (OR = 0.40 [0.18-0.91]; p = 0.028) or a P4 below 115 nmol/l (OR = 0.10 [0.01-0.52]; p = 0.006). CONCLUSIONS: Our study suggests a positive association between early P4 levels and reproductive outcomes in IVF using oral dydrogesterone for luteal support. The inconsistencies between our results and those of other studies suggest that extrapolation is impractical. Further larger prospective cohort studies should be conducted to determine reliable thresholds that could be used to personalize luteal phase support.