Where the numbers come from

Almost every fertility clinic in the United States reports its IVF results each year. Those reports are gathered by two public sources.

• SART, the Society for Assisted Reproductive Technology. Its member clinics report their cycles, patient ages and outcomes. A cycle is one round of IVF, from the fertility medicine to the egg retrieval.
• The CDC, the Centers for Disease Control and Prevention. It publishes clinic level results for clinics across the country. This data is in the public domain.

Our estimates are built from the cycles clinics report to SART. The CDC reported success rate for each clinic is shown in the clinic charts and used to check our work, but it is not used to build the estimates. Neither the CDC nor SART endorses this site.

We use results through 2024 and give estimates for 2026. We update the numbers when new clinic results come out. A few settings come from published studies rather than from our own data: how chances fall with age, the share of embryos that test normal, and the third transfer assumption described below.

Each clinic page can show two charts. Over time is our estimate of the chance of a baby for each year, with its range. Past years are fitted to what the clinic reported, which is not a test of our estimates. The dashed part is the forecast for 2025 and 2026. CDC reported shows the CDC reported success rate: the results the clinic itself reported to the CDC.

How we make an estimate

A clinic's past results alone can mislead. A clinic that treated 40 patients last year can look great or poor by luck. A clinic that mostly treats younger patients will look better than one that treats older patients, even if the care is the same.

So we do four things.

1. We compare like with like. Our method adjusts for your exact age and for diagnosis, so a clinic that treats older patients is not marked down for it.
2. We follow a patient through each step. Getting a baby takes several steps. First comes an egg retrieval, then embryos, then a transfer that leads to a birth. We estimate how many embryos a retrieval gives, how many survive freezing and, when embryos are tested, how many are chromosomally normal. We estimate each step on its own and then put them together.
3. We admit when we are unsure. A small clinic's own results count for less, so its estimate leans toward the national average and carries a wider range.
4. We let results drift over time. Each clinic's results can change from year to year. We carry that drift forward to 2026.

The result is an estimate for 2026. It is not a count of what the clinic already reported.

What chance of a baby means

No published data reports success rates for a third transfer on its own, so we assume it works as well as the second transfer (one embryo). Treat the 3 transfer number as a rougher guide.

We also show two smaller numbers.

• Chance from the first transfer. The chance of a baby from the first transfer, counted from the start of the cycle. It is lower than a clinic's per-transfer rate because it includes getting to a transfer.
• Chance of a usable embryo. The chance of getting at least one embryo that survives freezing and can be transferred. If all embryos are frozen first, it is the chance at least one survives thawing and, with PGT, tests normal. This is not a chance of a baby.

These are estimates. They are not the clinic's reported rates, and they are not a promise. Changing your age, PGT, diagnosis, egg source or gestational carrier can change the numbers.

We are least sure for ages 43 and over, patients who do not use PGT at 41 and over, third transfers, and clinics with few patients.

Uncertainty ranges

Next to each number you will see a 90% range. A narrow range means we have plenty of evidence. A wide range means we have less.

The range shows how much we are unsure about this clinic. On a clinic page, the range leaves out our uncertainty about how many embryos a cycle produces. The range on a clinic card includes it, so card ranges are wider. Neither covers every source of error, so treat it as a guide, not a guarantee. At older ages our numbers can run high, as the next section explains.

When two clinics have different numbers but their ranges overlap a lot, the difference may not be real. Please do not choose a clinic because of a gap of a few points.

Ages 41 and over

From 41 we count the first transfer, and show a second in a lighter colour. At these ages our numbers can run high, especially without genetic testing (PGT). When we checked a first cycle at 43 and over against 2023 and 2024 results, our estimate was about 15 percent too high. At 41 to 42 it was close. For a first transfer without PGT at 43 and over, our number was almost twice what actually happened. Treat these numbers as an upper limit.

Donor eggs and donor embryos

Donor estimates are for a cycle that uses eggs or embryos from a donor instead of your own. The starting choice is a frozen embryo that was made from donor eggs and is thawed for transfer, with PGT on (the same starting choice as for your own eggs) and no gestational carrier. Each donor cycle is separate. With donor eggs, the cycle has its own eggs and its own transfer. With donor embryos, the embryos already exist, so there is no egg step. We show one donor cycle, with one transfer. If a clinic reported no cycles with that kind of donation, it shows no estimate.

Embryos to transfer or freeze

Not every embryo is transferred right away. Embryos that are not used in a first transfer can be frozen for a later one. That is why the headline number can count more than one transfer from one cycle. How many we count depends on age, as described above. The chance of a usable embryo tells you how likely it is that a cycle gives you at least one embryo to transfer or freeze at all. Your doctor will talk with you about how many embryos to transfer at once and how many to freeze.

Prices and estimates

Where we can, we show what a clinic charges for one IVF cycle, and where the price comes from.

• Published or quoted prices come from the clinic's own website or price list, a quote from the clinic, the hospital's price file, or the clinic network's price. Some are part of a bundle price.
• Patient reports. Some prices are collected from patient reports online.
• Estimates. If none of these exist, we estimate from prices in the same region, or nationally. We mark it "estimate" and show a wider range: the real price falls inside it about 4 times in 5. Estimates are not a quote.
• Prices are shown without ICSI unless chosen. ICSI is intracytoplasmic sperm injection, a step where one sperm is placed directly into an egg. We leave it out of the total unless you choose to include it. It changes the price only.

We use prices dated within the last 6 years. Older prices are kept but not used, except genetic testing lab quotes, which show the year they were quoted. Prices change and often leave out medicines, testing and freezing. Always ask the clinic for a written quote.

Closed and merged clinics

Clinics close, move and join other clinics. We handle this in three ways.

• Closed clinics are removed from search and lists. If you open one, you will see that it has closed, and when we know them, the date, our source and which clinic it joined. No estimate is shown.
• Merged clinics send you to the page of the clinic they joined. If that page is not available, the old page links to continue care.
• Clinics with no estimate say "No estimate". This happens when the clinic reported no recent IVF data or no cycles like the ones you chose. We never fill the gap with a made-up number or a 0%. You can tap to see the US average beside it. It is labelled as the US average, not the clinic's.