IVF success after three cycles depends on how success is counted—and on the age of the woman starting treatment. A study of 18,396 women reported an optimal estimated cumulative live birth rate of 68.2%, but the rate among everyone who started treatment was 58.7%. These are cohort results, not an individual prediction.
The age-specific optimal estimate was 84.5% below age 35 and 57.7% at ages 38–40.
You have undergone one round of in vitro fertilization (IVF), but it did not lead to pregnancy. After egg retrieval, fertilization, culture, and transfer, the first questions are often: “Is there a chance if I try again? How much would three attempts change things?”
In fertility treatment, one measure is as important as the success rate of a single attempt: the cumulative live birth rate. It looks beyond the first treatment to the proportion of women who ultimately have a baby after continuing to a second or third treatment if the first did not lead to birth. Our explanation of why cumulative birth rates differ from single-cycle success rates explores this distinction.
A large study presented at the 2026 annual meeting of the European Society of Human Reproduction and Embryology (ESHRE) examined this question. Researchers analyzed up to three treatment cycles in 18,396 women who began their first IVF cycle between 2012 and 2021 at seven Australian fertility clinics, following birth outcomes through the end of 2023.
Why were the three-cycle estimates 68.2% and 58.7%?
The figure that stands out is 68.2%: the study’s optimal estimated cumulative live birth rate over up to three IVF cycles.
It should not be read as a promise that “seven out of ten people will have a baby if they try IVF three times.” The result changes depending on how people who discontinue treatment are handled.
When the calculation included all women who started treatment, the three-cycle cumulative live birth rate was 58.7%. The optimal statistical estimate, which accounts differently for incomplete follow-up and treatment continuation, was 68.2%.
The higher figure therefore describes how the chance of a birth may accumulate in this study population across continued treatment. It is not automatically “my success rate” for every patient facing infertility.
In this study, a cycle included an egg retrieval and the associated embryo transfers. Three cycles should not be confused with three individual embryo transfers.
How did cumulative live birth vary by age?
Age was a particularly important part of the result. The optimal estimated cumulative live birth rates over three cycles were:
- Under 35: 84.5%
- Ages 35–37: 74.4%
- Ages 38–40: 57.7%
- Ages 41–42: 30.1%
This is why the question “How many IVF attempts will it take?” cannot be answered by counting attempts alone. Even with the same three cycles, cumulative birth outcomes varied substantially by age at the start of treatment.
Individual circumstances also matter: ovarian function, the number of eggs retrieved, the number and condition of embryos created, and the cause of infertility. A success percentage found online cannot simply be adopted as a personal probability.
Most transfers used just one embryo
Another important feature was that 95.3% of all transfers were single embryo transfers. Despite transferring only one embryo at a time, the cohort achieved the reported cumulative birth results over three cycles, with a multiple birth rate of 2.9%.
In the past, transferring several embryos at once was relatively common as a way to increase the chance of pregnancy. Twin and higher-order multiple pregnancies, however, can increase risks for both mother and babies, including preterm birth and low birth weight.
The study illustrates a shift in contemporary IVF: from trying to succeed by transferring several embryos together, toward single embryo transfer with a chance of birth that accumulates across the course of treatment. The low multiple birth rate does not by itself establish long-term health outcomes.
One failed treatment does not determine every later outcome
The study’s message is not confined to the number 68.2%. A failed first embryo transfer does not determine the outcome of every subsequent treatment.
Frozen embryos from the first retrieval may be transferred later. If needed, another retrieval can begin a new treatment cycle. The chance of birth is considered across those steps, rather than through a single attempt alone. Evidence about sequential euploid embryo transfers after earlier failures addresses a related question, but uses a different population and unit of treatment.
This does not mean “three attempts are all you need.” The mean age in this study was 36.2 years, some patients used PGT-A, and the results came from seven clinics within one Australian network. The retrospective design and differences in who continued treatment also limit how widely the estimates can be applied.
The useful question in IVF is therefore broader than “What is my chance this time?” It also asks: “Given my age and ovarian condition, how might the chance of eventually having a child change if I continue treatment?”
That is the difference between a single-attempt success rate and the possibility of success across a complete course of treatment.
Study source and limitations
This article is based on D. Morbeck and colleagues’ “Modern IVF clinical practices achieve superior cumulative live birth rates with near-universal single embryo transfer: A multi-cycle cohort study,” published in Human Reproduction, volume 41, Supplement 1, in July 2026. The ESHRE conference abstract describes a retrospective cohort of 18,396 women starting IVF at seven Australian fertility clinics. DOI: 10.1093/humrep/deag083.327.
This is conference-abstract evidence, not an individual prognosis. The study did not assess long-term obstetric or neonatal outcomes.
Image disclosure: The image used in this article was created using generative artificial intelligence (ChatGPT, OpenAI). Any people depicted are fictional, not real individuals.
Copyright © 2026 Fertility Information TokTok · Fertility News | THE FERTILITY NOTES. All rights reserved. Unauthorized republication, reproduction, redistribution, scraping, AI training, and data use of any content, including articles, photographs, and images, are strictly prohibited. Violations may result in legal liability.