According to SART, for someone using their own eggs under 35 years old, one embryo transfer results in a live birth about 39.4% of the time. At 38 to 40 it’s 21.3%. Over 42 it’s 3.7%.
That same report also says for someone under 35, IVF works 68.2% of the time. At 38 to 40 it’s 43.2%. Over 42 it’s 7.5%.
Numbers all from SART, in 2023, describing the same patients. And they’re all accurate. They’re just describing different measures of success.
Same patients, four answers
SART is the professional body most US fertility clinics belong to, and their national report is the closest thing to an IVF scoreboard in America. They publish success rates four ways.
Live birth rates (with patient’s own eggs)
| Age | Per new patient | Per intendedretrieval | Per retrievalall transfers | Per retrievalfirst transfer |
|---|---|---|---|---|
| Under 35 | 68.2% | 42.8% | 53.2% | 39.4% |
| 35 to 37 | 58.1% | 30.5% | 39.9% | 31.0% |
| 38 to 40 | 43.2% | 19.4% | 26.2% | 21.3% |
| 41 to 42 | 24.1% | 9.4% | 13.2% | 11.3% |
| Over 42 | 7.5% | 2.8% | 4.1% | 3.7% |
SART Final National Summary Report, 2023. Live birth rates, patient’s own eggs. Every column describes the same patients in the same year. Only the denominator changes.
- Per new patient counts everything a person did across the year. Every retrieval, every transfer.
It answers: what are my chances… if I commit to IVF? - Per intended retrieval counts from the moment a retrieval cycle starts, including cancelled cycles or ones that never produce a transferable embryo.
It answers: what are my chances… from one retrieval? - Per retrieval, all transfers takes one egg retrieval and counts every embryo transfer that came out of it within that year.
It answers: what are my chances… from one retrieval, using the embryos it produces? - Per retrieval, first transfer counts only the first attempt. If that one fails and a later frozen embryo works, this number doesn’t see it.
It answers: what are my chances… from one retrieval, on the very first transfer?
So the question isn’t “what’s your success rate.”
It’s “per what?”
We go through how to read a clinic’s report on How to Choose an IVF Clinic, and where to find the CDC and SART databases on Comparing US Clinics.
How many rounds does it take?
For patients where IVF worked, here’s the average number of retrievals and transfers it took for them to bring home a baby.
| Age | Average egg retrievals | Average embryo transfers |
|---|---|---|
| Under 35 | 1.23 | 1.42 |
| 35 to 37 | 1.40 | 1.39 |
| 38 to 40 | 1.58 | 1.32 |
| 41 to 42 | 1.81 | 1.23 |
| Over 42 | 1.92 | 1.17 |
SART, 2023. Among patients who achieved a live birth, the mean number of egg retrievals and embryo transfers it took. This counts only the people it eventually worked for.
One caution. These averages only include people who ended up with a baby. For anyone where IVF didn’t work, they are not included. So the table can’t tell you how many rounds it takes to succeed, only how many it took for the people who did.
The transfer column runs the opposite direction from the retrieval column. Older patients needed more retrievals but fewer transfers because each retrieval yields fewer embryos, so there are fewer to transfer before they have to go back for another round.
Donor eggs and a bonus find
SART reports donor egg outcomes by cycle type.
| Donor cycle type | Live birth | Counted from |
|---|---|---|
| Fresh donor eggsRetrieved, fertilized, transferred fresh | 38.5% | Before embryos exist |
| Frozen donor eggsThawed, then fertilized | 37.9% | Before embryos exist |
| Thawed embryosDonor-egg embryos, already made and frozen | 46.6% | After embryos are made |
SART Final National Summary Report, 2023. Note the middle column. The first two count from a point where no embryo exists yet. The third counts after the embryos already exist. It is not a like-for-like comparison.
Fresh or frozen doesn’t matter. The two egg rows are nearly identical, and the research agrees. The cleanest test compared donor eggs transferred into gestational carriers, holding the uterus constant to isolate the freezing itself. No significant difference found.
That 46.6% for thawed embryos isn’t what it looks like. It’s not necessarily the “better” option because it’s not an equal comparison. The first two rows starts with healthy donor eggs, when no embryos existed. The third row starts with healthy embryos already made and frozen. It skips the fertilization and growing steps, and the weaker ones have already been teased out. It is a higher success rate, just counting from a different starting point.
Bonus find: It’s not you. It’s your eggs. SART’s report doesn’t include a column for age. It’s not missing by oversight. It’s missing because recipient age isn’t what drives the results. A UK study of 156,947 women further backed this up and concluded no age differential was observed among women using donor eggs. With donor eggs, a woman in her mid-forties and a woman in her thirties land in roughly the same place, around 40%.
This tells us what the age curve was measuring all along. It’s not the uterus. It’s the eggs.
The age curve isn’t about your body.
We all know as age goes up, success rates drop. But when (younger) donor eggs are used, age did not affect success.
The age curve is not about your uterus, your health, or your ability to carry a pregnancy. It’s the eggs. We can’t change someone’s eggs, but their eggs can be substituted.
Whether or not donor eggs is right for you is a big decision. But the numbers do point somewhere useful. It isn’t that IVF stops working as you get older. It’s that IVF with your own eggs gets harder.
What the numbers don’t know
They don’t know you.
Every figure here is an average across hundreds of thousands of people, all with different diagnoses, ovarian reserve, and histories. A 36-year-old with unexplained infertility and a 36-year-old with severe endometriosis are both categorized the same with the same published odds. But their real odds are nowhere near alike.
SART maintains a free patient predictor that asks about your specifics and estimates from there. It still won’t be your number, but it’s closer than a national average, and it’s a better thing to bring to a consultation than a statistic you read on a website. Including this one.
Sources
- Society for Assisted Reproductive Technology, “Final National Summary Report, 2023” — all rates on this page
- JAMA, “Live-birth rate associated with repeat in vitro fertilisation treatment cycles” — 156,947 women, donor egg age findings
- Centers for Disease Control and Prevention, “How to Interpret ART Success Rates”
- Society for Assisted Reproductive Technology, “IVF Success Predictor”
- Ginsburg et al., “Similar pregnancy outcomes from fresh and frozen donor oocytes transferred to gestational carriers: a SART database analysis isolating the effects of oocyte vitrification”