Progesterone-based ovulation timing versus LH surge monitoring in true natural cycle frozen embryo transfer: impact on live birth rates in a prospective cohort study.

ARTs have evolved substantially in recent decades, with frozen embryo transfer (FET) becoming increasingly used in clinical practice ( Roque, 2015 ). This shift has been largely driven by improvements in cryopreservation techniques and scheduling flexibility that FET offers, both of which may contribute to improved clinical outcomes ( Mackens et al. , 2017 ). Among FET protocols, true natural cycle frozen-thawed embryo transfer (NC-FET) has gained popularity because it preserves endogenous hormonal dynamics and avoids exogenous hormone administration ( Groenewoud et al. , 2013 ; Lawrenz et al. , 2020 ). In regularly menstruating women, accurate identification of ovulation requires close monitoring using serial ultrasound and hormonal measurements ( Erden et al. , 2022 ). Precise ovulation timing is essential in NC-FET to ensure optimal embryo-endometrium synchrony ( Aydin et al. , 2023 ). Serum LH monitoring is traditionally used to determine ovulation timing ( Irani et al. , 2017 ). Nevertheless, there is no universally accepted definition of the LH surge. Reported threshold values vary considerably across studies (e.g. ≥10, ≥15, ≥17 or ≥20 IU/l), and some studies combine LH measurements with estradiol levels or follicular collapse to confirm ovulation ( Groenewoud et al. , 2012 ; Irani et al. , 2017 ; Bartels et al. , 2019 ; Kahraman and Sahin, 2020 ; Erden et al. , 2022 ). This heterogeneity in defining the surge may lead to variation in the reference day used for scheduling embryo transfer. As a result, transfer timing may vary across protocols despite similar endocrine patterns. Such variability may result in biologically different luteal phase exposure at the time of embryo transfer. Moreover, the transient nature of the LH peak may introduce variability in scheduling embryo transfer, potentially affecting cycle reproducibility ( Testart et al. , 1981 ; Groenewoud et al. , 2012 ; Bartels et al. , 2019 ). Given the variability in LH-based definitions of ovulation, serum progesterone may provide a more direct endocrine marker of luteal transition, as progesterone exposure initiates secretory endometrial transformation and contributes to preparation of the implantation window ( Hsueh et al. , 2023 ). Progesterone levels increase at the time of ovulation and continue to rise in a progressive and sustained manner during the early luteal phase. In our previous prospective observations in subfertile women, we demonstrated a consistent peri-ovulatory rise in serum progesterone following ovulation, exhibiting a stable upward trajectory ( Aydin et al. , 2023 ). Similarly, Coughlan et al. (2023) defined ovulation using a progesterone threshold (≥1 ng/ml) and reported temporal discrepancies between LH surge detection and progesterone elevation. These findings highlight interindividual variability in endocrine sequencing around ovulation and suggest that LH-based and progesterone-based definitions may not identify the same reference day in all women. However, whether these temporal differences translate into clinically meaningful differences in embryo transfer outcomes remains unclear. To date, no prospective study has directly compared live birth outcomes between progesterone-timed and conventional LH-timed scheduling strategies in NC-FET cycles. Based on these considerations, we conducted a prospective comparison to determine whether progesterone-timed scheduling yields live birth rates comparable to conventional LH surge timing in true NC-FET.

A total of 490 NC-FET cycles were analyzed, including 286 P4-timed and 204 LH-timed cycles. Baseline demographic and clinical characteristics are summarized in Table 1 . Baseline characteristics and obstetric outcomes according to ovulation timing method. Continuous variables are presented as mean ± SD or median (interquartile range, IQR) according to distribution normality. Categorical variables are presented as number (percentage). Comparisons between the P4-Timed and LH-Timed groups were performed using Student’s t -test or Mann–Whitney U- test for continuous variables, and χ 2 test for categorical variables. A two-sided P  0.10 were considered indicative of meaningful imbalance. SMDs and P -value for categorical variables were reported for the overall variable. ET, embryo transfer; PGT, preimplantation genetic testing. The two groups were generally well balanced with respect to age, infertility duration, previous ART attempts, cycle length, endometrial thickness at transfer, embryo transfer type, embryo quality, and infertility etiology. Mean maternal age was 31.5 years in both groups. Body mass index was slightly higher in the P4-timed group (24.70 ± 4.53 kg/m 2 vs 23.98 ± 4.17 kg/m 2 ; P  = 0.075), although the SMD suggested only modest imbalance (SMD = 0.163). Most other baseline and treatment-related variables showed small between-group differences, although previous pregnancy loss and ET-day progesterone differed statistically between groups. PGT was performed in 57 of 490 cycles (11.6%). Its use was more frequent in the LH-timed group than in the P4-timed group (37/204, 18.1% vs 20/286, 7.0%; P  < 0.001). Among cycles resulting in live birth, obstetric outcomes were similar between groups. Mean gestational age at delivery was 37.92 ± 3.34 weeks in the P4-timed group and 37.90 ± 3.55 weeks in the LH-timed group ( P  = 0.965). Singleton deliveries accounted for 93.8% and 87.7% of births, respectively, while twin births occurred in 6.2% and 12.3% of cases ( P  = 0.117). Neonatal birth weight was also comparable between groups. Mean singleton birth weight was 3258.7 ± 510.4 g in the P4-timed group and 3321.8 ± 437.0 g in the LH-timed group ( P  = 0.325). Among twin deliveries, mean birth weight was 2290.5 ± 452.7 g and 2416.7 ± 402.1 g, respectively ( P  = 0.487) ( Table 1 ). Primary and secondary reproductive outcomes are presented in Table 2 . Live birth occurred in 162 of 286 P4-timed cycles (56.6%) and 106 of 204 LH-timed cycles (52.0%), corresponding to an unadjusted RR of 1.09 (95% CI 0.93–1.28; P  = 0.295). In the adjusted modified Poisson regression model, ovulation timing strategy was not significantly associated with live birth (aRR 1.09, 95% CI 0.93–1.30; P  = 0.264). Adjusted live birth probabilities derived through marginal standardization were ∼57.2% for P4-timed cycles and 52.3% for LH-timed cycles, yielding an adjusted absolute RD of +4.9% (95% CI −3.8% to +13.6%) ( Fig. 1 ). In an exploratory threshold-based assessment, the lower bound of the adjusted RD for live birth was −3.8%, which did not cross the candidate absolute risk-difference thresholds of −10.0%, −7.5%, or −5.0%. This analysis was interpreted descriptively and was not considered evidence of formal non-inferiority ( Supplementary Table S1 ). Adjusted effects of ovulation timing strategy on reproductive outcomes . ( A ) Adjusted relative risks (aRRs) with 95% CIs for live birth, clinical pregnancy, hCG positivity, and miscarriage comparing progesterone-timed (P4-timed) and luteinizing hormone-timed (LH-timed) groups. ( B ) Adjusted absolute risk differences (RDs) with 95% CIs (P4 − LH). Primary and secondary outcomes by ovulation timing strategy with adjusted risk estimates and absolute risk differences. Unadjusted relative risks (RRs) with 95% CIs were calculated for binary outcomes. Adjusted relative risks (aRRs) were estimated using modified Poisson regression with robust standard errors, adjusting for maternal age, BMI, infertility duration, infertility etiology, endometrial thickness on the day of transfer, PGT status, number of embryos transferred, and embryo quality. Adjusted absolute risks were derived from the primary model using marginal standardization. Adjusted risk differences (RDs) and corresponding 95% CIs were calculated using the delta method. Secondary outcomes were also comparable between groups. Clinical pregnancy occurred in 67.5% of P4-timed cycles and 62.3% of LH-timed cycles (unadjusted RR 1.08; 95% CI 0.94–1.23; P  = 0.219). After adjustment, the aRR was 1.09 (95% CI 0.96–1.24; P  = 0.167) with an adjusted RD of +5.9% (95% CI −2.4% to +14.1%). hCG positivity occurred in 74.1% versus 70.6% of cycles (adjusted aRR 1.07; 95% CI 0.96–1.19; P  = 0.207). Miscarriage rates among clinical pregnancies were similar (16.1% vs 16.5%; adjusted aRR 1.01; 95% CI 0.61–1.67; P  = 0.958) ( Table 2 ). Multivariable adjusted analyses of live birth are presented in Table 3 . In the primary modified Poisson regression model, the ovulation timing method was not significantly associated with live birth (aRR 1.07; 95% CI 0.89–1.28; P  = 0.47). Higher progesterone concentrations on the day of embryo transfer were modestly associated with an increased probability of live birth (aRR 1.02 per 1 ng/ml increase; 95% CI 1.00–1.04; P  = 0.041). Infertility duration remained inversely associated with live birth probability (aRR 0.97 per year; 95% CI 0.94–0.99; P  = 0.021). Double embryo transfer was independently associated with a higher probability of live birth compared with single embryo transfer (aRR 1.38; 95% CI 1.16–1.64; P  < 0.001). Other covariates, including maternal age, BMI, infertility etiology, endometrial thickness, and embryo quality, were not significantly associated with live birth. Results from the multivariable logistic regression sensitivity analysis were consistent with the primary model (aOR 1.16; 95% CI 0.76–1.78; P  = 0.49), supporting the robustness of the findings. Multivariable adjusted analyses of live birth according to ovulation timing method; primary modified Poisson regression and logistic regression sensitivity analysis. aRR derived from modified Poisson regression with robust standard errors; aOR derived from multivariable logistic regression (sensitivity analysis). Continuous variables modeled per unit increase. Hormonal and follicular parameters according to ovulation timing method are presented in Table 4 . Baseline hormonal profiles at menses were comparable between groups. During the pre-ovulatory phase, follicle size measured 1 day prior to ovulation was slightly larger in the LH-timed group, while estradiol and progesterone levels were similar between groups. Hormonal and follicular parameters according to ovulation timing method. Sample sizes vary by hormonal time point, as repeated early luteal measurements were obtained as clinically indicated rather than mandated daily. In P4-timed cycles, luteal day was assigned at the first qualifying progesterone measurement. During the peri-ovulatory and early luteal phases, no significant differences were observed in estradiol or progesterone concentrations on the day of ovulation or in the days immediately following ovulation. However, progesterone levels on the day of embryo transfer were modestly higher in the LH-timed group compared to the progesterone-timed group. Hormonal relationships around ovulation and embryo transfer were explored using scatter plots with regression lines ( Fig. 2 ). ET-day progesterone demonstrated a moderate positive correlation with follicle size measured 1 day prior to ovulation ( r  = 0.36, P  < 0.001). A positive correlation was also observed between ET-day progesterone and estradiol concentrations measured on OvU − 1 day ( r  = 0.29, P  < 0.001). Hormonal relationships around ovulation and embryo transfer in natural-cycle frozen embryo transfer cycles . ( A ) Relationship between follicle size measured 1 day before ovulation (OvU-1) and progesterone concentration on the embryo transfer (ET) day. ( B ) Relationship between progesterone concentration on the embryo transfer day and estradiol levels measured 1 day before ovulation (OvU-1). Regression lines with 95% CIs are shown separately for progesterone-timed (P4-timed) and luteinizing hormone-timed (LH-timed) cycles. Overall, reproductive outcomes were comparable between the P4-timed and LH-timed strategies, with no evidence that ovulation timing method was independently associated with live birth.

This prospective cohort study was conducted at the In Vitro Fertilization Unit of Acibadem Mehmet Ali Aydinlar University, Atakent Hospital (Istanbul, Turkey), between October 2022 and April 2025. The study protocol was approved by Acibadem Mehmet Ali Aydinlar University ethics committee (ATADEK-2022/15), and written informed consent was obtained from all participants prior to inclusion. All procedures were performed in accordance with the ethical principles of the Declaration of Helsinki ( Shrestha and Dunn, 2020 ) and Good Clinical Practice guidelines ( Handbook for Good Clinical Research Practice (GCP): Guidance for Implementation, 2005 ). A total of 518 women aged 18–40 years undergoing NC-FET were initially assessed for eligibility. Five patients were excluded due to progesterone levels below 4 ng/ml on the day of embryo transfer, which was considered inconsistent with adequate luteal progression for NC-FET scheduling. These cycles were therefore not included in the completed-transfer analysis. Additionally, 12 patients were excluded due to inadequate follicular development or discordant follicular and hormonal findings that prevented reliable assignment of the endocrine reference day, seven patients canceled the embryo transfer due to personal reasons, and in four patients, embryo transfer was not performed due to embryo degeneration and the absence of an alternative embryo for transfer. Inclusion criteria were as follows: Regular menstrual cycles ranging from 24 to 38 days, Normal ovarian morphology and baseline endocrine profile, and Availability for serial hormonal and ultrasonographic monitoring. Patients were excluded if they exhibited: Anovulatory polycystic ovary syndrome or other endocrine abnormalities (thyroid disease, hyperprolactinemia); Baseline progesterone >1.0 ng/ml on cycle Days 2–7; Inability to comply with the monitoring protocol; or Uterine malformations, untreated intrauterine pathology, or untreated/surgically uncorrected hydrosalpinx. Tubal factor infertility was not an exclusion criterion if hydrosalpinx was absent or had been surgically treated before embryo transfer. Baseline evaluation was performed by transvaginal ultrasonography (GE Voluson S8, Austria) between cycle Days 2 and 7 to exclude ovarian cysts and assess antral follicle count. Beginning from cycle Days 8 to 13 onward, patients underwent serial ultrasonography every 2–3 days until the dominant follicle reached 15–16 mm in diameter. At this point, daily or alternate-day hormonal monitoring was initiated, including measurements of serum E2, LH, and progesterone levels. No hCG trigger was used, and ovulation occurred spontaneously in all cycles. Hormonal monitoring was continued until an endocrine reference day for embryo transfer scheduling could be assigned; daily progesterone measurement until embryo transfer was not mandated for all cycles. The ovulation timing strategy was not randomized and was not prespecified at the individual patient level before cycle initiation. Once the endocrine reference day was identified and embryo transfer was scheduled, the timing strategy was not changed. Two ovulation monitoring strategies were used: LH-timed group: Ovulation was defined as the day serum LH concentration reached ≥15 IU/l ( Ursillo et al. , 2021 ) and with a concomitant progesterone level <1.0 ng/ml (LH surge day, LH+0). Blastocyst transfer was scheduled on LH+6. Ovulation was primarily defined by the LH surge, and transvaginal ultrasonography was used to confirm follicular collapse when needed. In the progesterone-timed (P) group, embryo transfer scheduling was based on serum progesterone concentrations interpreted according to the algorithm derived from our previous prospective ovulation-monitoring study ( Aydin et al. , 2023 ). The same institutional laboratory platform was used in the previous ovulation-monitoring cohort and in the present study; therefore, the thresholds were applied within the same assay environment. In that study, serial hormone measurements showed a predictable rise in progesterone in relation to the day of ovulation, supporting its use as a marker of early luteal transition. In the present study, the progesterone thresholds were used to assign the estimated luteal day at the time of the first qualifying progesterone measurement, rather than to reclassify the cycle repeatedly on subsequent days. The first serum progesterone value ≥1.0 ng/ml was considered to indicate the onset of luteal transformation. At the time of first detection, progesterone concentrations of 1.0–1.6 ng/ml were classified as estimated luteal Day 0, 1.6–2.6 ng/ml as estimated luteal Day 1, and 2.6–5.0 ng/ml as estimated luteal Day 2. Blastocyst transfer was then scheduled for estimated luteal Day 5; therefore, transfer was scheduled 5 days after detection for cycles classified as Day 0, 4 days after detection for cycles classified as Day 1, and 3 days after detection for cycles classified as Day 2. Subsequent progesterone measurements, when obtained, were used as supportive information and did not reset the assigned reference day. Cycles with discordant follicular and hormonal findings that prevented reliable luteal-day assignment were excluded. In some cycles, ovulation had already occurred by the time of the scheduled ultrasound assessment, and a preovulatory dominant follicle was no longer visualized. These cycles were not canceled. Instead, patients were assigned to the progesterone-based timing group according to the first qualifying serum progesterone level and the corresponding estimated luteal day, and embryo transfer timing was scheduled accordingly. Blastocysts were vitrified on Day 5 or 6 using Kitazato Vitrification Kit (BioPharma, Shizuoka, Japan) and graded according to the Gardner and Schoolcraft system ( Gardner and Schoolcraft, 1999 ). Preimplantation genetic testing (PGT) status was recorded for all cycles and was included in the baseline comparison between groups. Warming was performed using the Kitazato Warming Kit (BioPharma). Post-warming, embryos were cultured for ∼2 h to confirm complete re-expansion before transfer. One or two vitrified-warmed blastocysts were transferred under transabdominal ultrasound guidance using a soft Wallace catheter (Smiths Medical, UK) by two experienced physicians to minimize procedural variability. The number of embryos transferred was determined based on female age, prior treatment history, and shared decision-making. Luteal support was initiated immediately after transfer with oral dydrogesterone (30 mg/day) and continued until serum β-hCG measurement 12 days later. In cases of confirmed pregnancy, luteal support was maintained until 6–8 weeks of gestation. All hormonal measurements were performed at the Acıbadem Mehmet Ali Aydınlar University Atakent Hospital Central Biochemistry Laboratory using a chemiluminescent enzyme immunoassay (Atellica IM Analyzer, Siemens Healthineers, Germany). Analytical performance characteristics were as follows: Progesterone: sensitivity ≤0.21 ng/ml; intra-assay CV ≤13% for 1.0–6.5 ng/ml, ≤6% for 6.6–20 ng/ml, ≤6.5% for >20 ng/ml. LH: sensitivity ≤0.2 IU/l; CV ≤6.5%. Each serum sample was analyzed in duplicate within 2 h of collection to minimize analytical variability. The primary outcome was live birth rate, defined as delivery of a live infant after 24 weeks of gestation. Secondary outcomes included clinical pregnancy rate (presence of gestational sac with fetal heartbeat), hCG positivity (transient β-hCG rise without ultrasound confirmation), miscarriage rate (pregnancy loss before 24 weeks), gestational age at delivery, and neonatal birth weight. Baseline characteristics were summarized as mean ± SD for normally distributed continuous variables and as median with interquartile range (IQR) for skewed variables. Categorical variables were presented as counts and percentages. Between-group comparisons were performed using Student’s t -test for normally distributed continuous variables, the Mann–Whitney U -test for non-normally distributed variables, and the χ 2 test or Fisher’s exact test for categorical variables, as appropriate. Standardized mean differences (SMDs) were calculated to quantify between-group imbalance independent of sample size; values >0.10 were considered indicative of meaningful imbalance. The primary outcome was live birth, analyzed as a binary endpoint (live birth vs all other outcomes). Secondary outcomes included clinical pregnancy, hCG positivity, and miscarriage. The primary exposure was ovulation timing method (progesterone-timed [P4-timed] vs LH-timed). For outcome comparisons, unadjusted relative risks (RRs) with 95% CIs were calculated. Adjusted relative risks (aRRs) were estimated using modified Poisson regression with robust standard errors. Covariates included maternal age, body mass index (BMI), infertility duration, infertility etiology, endometrial thickness on the day of transfer, number of embryos transferred, embryo quality, progesterone concentration on the embryo transfer day, and PGT status. Continuous variables were modeled per 1-unit increase. Adjusted absolute risks were derived from the primary modified Poisson model using marginal standardization. The adjusted absolute risk difference (RD) between P4-timed and LH-timed cycles was calculated. CIs for adjusted RDs were obtained using the delta method. As an exploratory threshold-based assessment, the lower bound of the adjusted absolute RD for live birth was compared with candidate absolute risk-difference thresholds of −10.0%, −7.5%, and −5.0%. These thresholds were selected to represent progressively stricter clinically interpretable differences in live birth rate. The −10.0% threshold was considered the largest acceptable absolute difference and is consistent with margins used in ART embryo-transfer non-inferiority trial protocols ( Neuhausser et al. , 2020 ). The −7.5% threshold was selected as an intermediate threshold informed by FET non-inferiority literature, in which a 7.5% minimal clinically important difference has been discussed ( Groenewoud et al. , 2016 ). The −5.0% threshold was included as a conservative sensitivity threshold ( Illingworth et al. , 2024 ). Because the present study was observational and was not powered using an a priori non-inferiority framework, this assessment was considered exploratory and was not used to establish formal non-inferiority. As a sensitivity analysis, multivariable logistic regression including the same covariates was performed to estimate adjusted odds ratios (aORs). Additionally, exploratory correlation analyses were performed to evaluate hormonal relationships around ovulation and embryo transfer. Spearman rank correlation coefficients (r) were calculated to assess associations between follicle size measured 1 day prior to ovulation (OvU − 1) and progesterone levels on the embryo transfer day, as well as between estradiol concentrations measured on OvU − 1 day and embryo transfer-day progesterone levels. All analyses were conducted using robust variance estimation, and statistical significance was defined as a two-sided P -value < 0.05. Analyses were performed using R version 4.5.1 (R Foundation for Statistical Computing, Vienna, Austria) and GraphPad Prism version 10.0 (GraphPad Software, San Diego, CA, USA).

In conclusion, determining embryo transfer timing using serum progesterone levels results in reproductive outcomes comparable to those achieved with LH-based scheduling in true NC-FET. These findings suggest that serum progesterone may be a feasible alternative marker for ovulation detection and embryo transfer scheduling in true NC-FET. Given the observational design, the findings should be interpreted as showing comparable outcomes and no evidence of a difference. Confirmation in adequately powered multicenter studies, ideally incorporating assay-specific validation or calibration of progesterone thresholds, is required before broader clinical implementation.

In this prospective cohort study including 490 NC-FET cycles, progesterone-based embryo transfer timing resulted in reproductive outcomes comparable to those achieved with conventional LH-based scheduling. Clinical pregnancy, implantation, miscarriage, and live birth rates were similar between groups, with no evidence of an independent association between ovulation timing strategy and live birth after adjustment for potential confounders. These findings suggest that serum progesterone may represent a clinically useful alternative marker for determining the optimal embryo transfer day in NC-FET cycles. Importantly, progesterone monitoring may reduce the risk of missing the LH surge and simplify cycle monitoring without compromising treatment success. Achieving optimal outcomes in NC-FET requires close alignment between embryo transfer timing and the period of endometrial receptivity. Although the LH surge has traditionally been used to define ovulation, its clinical interpretation remains variable. Some women exhibit a brief, single LH peak, whereas others display prolonged or biphasic surges ( Direito et al. , 2013 ), introducing uncertainty when estimating the exact timing of ovulation and selecting the optimal embryo transfer day. Previous studies have highlighted that the definition of the LH surge in NC-FET varies considerably across centers ( Erden et al. , 2022 ). For example, identifying the surge based on the first attainment of LH ≥ 17 IU/l combined with a ≥ 30% decline in estradiol has been associated with higher implantation and live birth rates compared with defining the surge as the highest LH value measured on the following day ( Irani et al. , 2017 ). Other studies evaluating vitrified-warmed transfers after an LH surge ≥ 20 IU/l reported similar outcomes when blastocyst transfer occurred 6 or 7 days after the surge, suggesting some clinical flexibility but also underscoring the lack of standardization in LH-based timing strategies ( Bartels et al. , 2019 ). When follicular collapse is used as the reference for ovulation, the timing of blastocyst transfer appears consistent; however, when these same cycles are retrospectively reclassified according to varying LH-based definitions, the estimated transfer window shifts considerably, ranging from LH + 6 and LH + 9 ( Erden et al. , 2022 ). Ultrasound confirmation of ovulation by follicular collapse also has important limitations. Follicular evacuation during human ovulation may occur in different patterns, and residual follicular fluid can persist even after ovulation and into the luteal phase ( Hanna et al. , 1994 ). Therefore, incomplete or delayed follicular collapse does not necessarily indicate absence of ovulation, and reliance on follicular collapse alone may lead to uncertainty in defining the ovulation reference day. In the present study, ultrasonographic follicular collapse was therefore used only as supportive confirmation when needed, rather than as the primary criterion for scheduling embryo transfer in the LH-timed group. Collectively, these findings demonstrate that LH-based criteria remain heterogeneous across clinical practice, limiting reproducibility and complicating standardization of NC-FET timing. Unlike LH, which primarily signals the trigger of ovulation, progesterone reflects the initiation of endometrial transformation and plays a key role in determining endometrial receptivity. This biological specificity may provide a more physiologically relevant reference point for defining the beginning of the luteal phase. Previous studies have suggested that progesterone rise, particularly beyond ∼1–1.5 ng/ml, may more closely reflect the opening of the implantation window than LH measurements alone and therefore serve as a practical marker for scheduling blastocyst transfer ( Ata et al. , 2023 ). However, emerging evidence indicates that the temporal relationship between the LH surge and progesterone rise is not always consistent. In a recent observational study, Coughlan et al. (2023) demonstrated substantial interindividual variation in the interval between LH elevation and subsequent progesterone increase in natural cycles. Notably, in that study, embryo transfers were scheduled exclusively based on progesterone, and comparative reproductive outcomes relative to LH surge-based timing were not evaluated ( Coughlan et al ., 2023 ). To date, direct comparative evidence between progesterone-based and LH surge-based embryo transfer timing in natural cycles remains limited. A retrospective study comparing these transfer strategies reported higher clinical pregnancy rates in the progesterone-based group ( Dong et al. , 2014 ). In this context, our prospective findings provide further evidence that progesterone-based timing was associated with outcomes comparable to LH-based scheduling. To further explore the physiological basis of progesterone-based timing, we performed exploratory hormonal analyses examining the relationships between follicular development and luteal hormone dynamics. We observed a moderate positive correlation between follicle size measured 1 day before ovulation and progesterone concentrations on the embryo transfer day, suggesting that luteal progesterone dynamics may reflect the preceding degree of follicular maturation. In addition, progesterone levels on the embryo transfer day were positively correlated with preovulatory estradiol concentrations, supporting a physiological link between preovulatory follicular activity and subsequent corpus luteum function. Taken together, these findings suggest that progesterone-based timing captures the underlying endocrine transition from follicular development to luteal phase more directly than LH surge detection alone, thereby offering a physiologically grounded and clinically relevant alternative to LH-based timing strategies. From a clinical standpoint, progesterone-based monitoring may offer practical advantages. It simplifies NC-FET follow-up by reducing dependence on repeated LH measurements and frequent follicular ultrasound scans, which can be logistically challenging and emotionally burdensome for patients. This approach may reduce the number of monitoring visits, decrease patient burden, and improve feasibility in resource-limited settings. Importantly, in LH-based strategies, failure to capture the LH surge may lead to cycle cancellation, whereas progesterone-based monitoring may reduce this risk by providing a more reliable indicator of luteal transition. In contrast, LH surge characteristics demonstrate substantial interindividual variability in both amplitude and duration, and there is no universally accepted LH threshold for defining ovulation in clinical practice. This heterogeneity limits the standardization of NC-FET timing when LH is used as the primary marker. Consequently, progesterone monitoring might allow clinicians to identify ovulation more confidently, particularly in cycles where LH dynamics are atypical, prolonged, or inconclusive. The comparable live birth rates observed between the two groups in our study support the clinical feasibility of progesterone-based timing in NC-FET cycles. Importantly, this similarity persisted after adjustment for multiple clinical and treatment-related variables, indicating that the ovulation-timing strategy itself was not an independent determinant of live birth. Consistent findings across both modified Poisson regression and logistic regression sensitivity analyses further strengthen the robustness of this observation. Among the evaluated covariates, higher progesterone concentrations on the day of embryo transfer and double embryo transfer were variables independently associated with a higher probability of live birth, whereas infertility duration remained inversely associated with live birth probability. Other factors, including maternal age, BMI, infertility etiology, endometrial thickness, and embryo quality, were not significant predictors after adjustment. Several limitations should be acknowledged. First, although the study was prospective and included a relatively large cohort, it was conducted at a single tertiary IVF center. Second, the ovulation timing strategy was not randomized and was not prespecified at the individual patient level before cycle initiation. Instead, group assignment was determined dynamically based on hormonal measurements obtained during monitoring. Cycles were classified as LH-based when the first serum LH value was ≥15 IU/l with a concomitant progesterone level <1.0 ng/ml. Cycles were classified as progesterone-based when serum progesterone reached ≥ 1.0 ng/ml, irrespective of LH levels. Although this threshold-based approach reflects real-world clinical decision-making, it may introduce selection bias despite adjustment for measured baseline and treatment-related variables. Third, the study was not designed or powered as a formal non-inferiority trial. The threshold-based risk-difference assessment should therefore be interpreted as exploratory and descriptive rather than as definitive evidence of clinical non-inferiority. Fourth, the progesterone thresholds used to estimate luteal day were derived from a previous single-center prospective cohort and applied using the same institutional assay environment. Progesterone immunoassays may differ in calibration, analytical sensitivity, precision in the low peri-ovulatory range, and antibody-related assay characteristics; therefore, these thresholds should be regarded as assay- and setting-specific until externally validated. Finally, cycles with serum progesterone <4 ng/ml on the scheduled embryo transfer day were canceled and were not included in the completed-transfer analysis. This may limit applicability to patients with inadequate luteal progression or luteal phase defects. Multicenter studies incorporating assay-specific validation or calibration of progesterone thresholds are needed before broader clinical implementation.