Abstract Study question Are live birth rates after natural (NC), modified natural (mNC) endometrial preparation superior to artificial cycle (AC) strategies for ovulatory women undergoing frozen embryo transfer? Summary answer Live birth rates were comparable between NC, mNC, and AC. However, NC and mNC resulted in a significantly higher cycle cancellation rate. What is known already The number of frozen embryo transfer (FET) cycles has increased dramatically. In ovulatory women undergoing FET, three common strategies can be used for endometrial preparation: NC, mNC (NC + human chorionic gonadotropin trigger), and AC protocols. Previous studies comparing the effectiveness of these strategies had several drawbacks, including small sample size, lack of consistency between studies in the assessment and reporting of outcomes, and the retrospective origin, which introduced potential biases. Therefore, there is a lack of consensus on the optimal endometrial preparation protocol for FET. Study design, size, duration This single-center, open-label, randomized controlled trial was conducted at IVFMD, My Duc Hospital in Ho Chi Minh City, Vietnam. From March 2021 to March 2022, a total number of 1,428 participants were randomized (476 per group). The primary outcome of the study was live birth after one FET. Secondary outcomes were fertility outcomes, early pregnancy complications, obstetrical outcomes, perinatal and neonatal complications, and endometrial preparation cycle cancellation rates. Participants/materials, setting, methods Eligible participants were ovulatory women aged 18–45 with FET treatment, underwent no more than three IVF/intracytoplasmic sperm injection-FET cycles, had no more than two day-3 or one blastocyst transferred. Women with menopause/anovulation, contraindications for hormonal administration, uterine abnormalities who underwent in-vitro maturation, pre-implantation genetic testing, oocyte donation cycles were excluded. The FET strategies included FET after the first endometrial preparation cycles with AC, mNC, or NC, as per randomization, and the second with AC. Main results and the role of chance A total of 4,779 participants were screened, of whom 1,428 eligible individuals (476 per group) were randomized after written informed consent. The live birth rate after one FET was 36.6% (174/476) in the natural cycle strategy group, 33.4% (159/476) in the modified natural cycle strategy group, and 34.0% (162/476) in the artificial cycle strategy group (relative risk [95% confidence interval] for natural versus artificial cycle strategy: 1.07 [0.87–1.33] and for modified natural versus artificial cycle strategy: 0.98 [0.79–1.22]). Maternal and neonatal outcomes were comparable between groups, in particular,hypertensive disorders in pregnancy. For both natural and modified cycle groups, 20.08% (99/476) of the first FET cycles were canceled before subsequent successful ET versus none in the artificial cycle group. Limitations, reasons for caution Due to the nature of the intervention, blinding was not possible. The study was conducted in ovulatory women in a single center in Vietnam, which may limit the generalisability of the findings. Wider implications of the findings Live birth rates and obstetric/neonatal outcomes after the natural cycle, modified natural cycle, or artificial cycle strategies for ovulatory women undergoing FET are comparable, with higher cancellation rates in both the natural cycle and the modified natural cycle groups. Trial registration number NCT04804020