Abstract Study question Does the live birth rate differ between the two regimens for luteal phase supplementation in IVF cycles in women with a history of IVF failure? Summary answer The live birth rate did not differ with dydrogesterone + progesterone gel vs. subcutaneous aqueous progesterone + progesterone gel. What is known already Supplementation of the luteal phase in in vitro fertilization (IVF) cycles is crucial to increase the chance of a live birth. Progesterone is the treatment of choice due to a lower risk of complications compared to human chorionic gonadotropin (hCG) and gonadotropin-releasing hormone (GnRH) analogues. There is no preferred progestogen for use in the general population of women undergoing IVF. The optimal progesterone treatment regimen in the event of previous IVF failure is unknown. Study design, size, duration A prospective randomized tertiary single-center study was conducted. The study included 161 women, aged 25-43 years, with FSH on cycle day 2-4 ≤10 IU/L, with at least one failed IVF cycle, undergoing government-funded IVF in the years 2015-2016. Participants/materials, setting, methods 161 included women were randomized after oocytes retrieval to 2 groups of luteal IVF cycle phase supplementation: dydrogesterone (Duphaston®) + progesterone in vaginal gel (Crinone®) or aqueous progesterone solution in subcutaneous injection (Prolutex®) + progesterone in vaginal gel (Crinone®). At least one blastocyst was obtained in the enrolled women. All women underwent fresh embryo transfer. After viable pregnancy confirmation, treatment was continued until the 12th week of gestation. Main results and the role of chance The indications for IVF were: male factor (51%), tubal factor (27%), endometriosis (24%), decreased ovarian reserve (17%), idiopathic infertility (14%), ovulation disorders (6%), occurring in various combinations. Out of 161 women enrolled in the study, 94 had a history of 1 failed IVF cycle, and 67 had at least 2 failed IVF cycles. Among 94 women with 1 failed IVF, 62 received dydrogesterone + progesterone gel vs. 32 aqueous progesterone + progesterone gel, and the percentages of live births, clinical pregnancies and biochemical pregnancies were: 29 vs. 21% (p = 0.37), 10 vs. 6% (p = 0.53), 5 vs. 12% (p = 0.21). Among 67 women with at least 2 failed IVFs, 24 received dydrogesterone + progesterone gel vs. 43 aqueous progesterone + progesterone gel, and the percentages of live births, clinical pregnancies and biochemical pregnancies were: 17 vs. 32% (p = 0.19), 4 vs. 0% (p = 0.14), 9 vs. 9% (p = 0.95). The results were comparable with both treatment regimens. There were no significant adverse events reported for both treatment regimens. Limitations, reasons for caution The limitations of the study are the small sample size, its single-center character and heterogeneity of indications for IVF. Wider implications of the findings In light of the lack of evidence of greater efficacy of either of the two luteal phase supplementation regimens of the IVF cycle in women with a history of IVF failure, the choice of treatment should be guided by other considerations, i.e. potential side effects, dosing convenience and patient preference. Trial registration number 1072.6120.47.2021