How to Read IVF Success Rates (Per Retrieval vs. Per Transfer)

A single fertility clinic can be described with very different “success rates” depending on what is counted. Before you compare clinics, it helps to know which number you are looking at. This guide explains the three main CDC measures and shows how much they differ, using the CDC’s 2022 national figures.^[2]

The short answer

• Per intended egg retrieval is the most complete measure for patients using their own eggs. It counts every cycle started with the goal of retrieving eggs, including cycles that never reached an embryo transfer.^[1]
• Per actual retrieval leaves out cycles where no eggs were retrieved.
• Per transfer counts only the cycles that reached an embryo transfer. It is usually the highest number, because earlier drop-offs are excluded.

FertilityList ranks clinics on live births per intended retrieval for this reason. It is the measure least affected by which patients reach the later stages.

What each measure counts

The CDC describes an ART cycle as starting when a patient begins fertility medication or ovarian monitoring with the plan to retrieve eggs. That is an intended retrieval. If at least one egg is collected, it becomes an actual retrieval. Embryos may then be transferred fresh or frozen for later use.^[1]

For patients using their own eggs, the CDC reports cumulative rates. These count live births from all embryo transfers that happen within one year after the retrieval, both fresh and frozen.^[3]

How much the numbers differ: 2022 national data

The table shows national live-birth rates for patients using their own eggs, by age group and by measure.^[2]

Age	Per intended retrieval	Per actual retrieval	Per transfer
Under 35	49.7%	52.6%	49.1%
35–37	36.3%	39.7%	46.0%
38–40	23.1%	26.1%	41.2%
Over 40	7.7%	9.4%	25.4%

The gap between the first and last columns widens with age. For patients over 40, the per-transfer rate is 25.4%, while the per-intended-retrieval rate is 7.7%.^[2] Many cycles in that group end before any embryo is transferred, so a per-transfer figure on its own can look far more optimistic than what a new patient experiences.

Why per-transfer rates can mislead

Imagine two clinics that each start 100 egg-retrieval cycles for patients over 40. Clinic A transfers embryos in 60 of those cycles; Clinic B transfers in only 30, perhaps because it cancels more cycles or tests embryos more strictly. If both end up with the same number of babies, Clinic B’s per-transfer rate will be roughly double Clinic A’s, even though a patient walking in the door had the same chance at either clinic. The per-intended-retrieval rate treats both clinics the same way, because it starts counting at the first step. This example uses round, illustrative numbers rather than real clinic data.

The same logic applies to clinics that freeze all embryos and transfer later, or that use preimplantation genetic testing (PGT) on most embryos. These practices change how many cycles reach a transfer, and so change the per-transfer rate, without necessarily changing the overall chance of a baby per started cycle. Each clinic page on FertilityList shows how often that clinic used frozen transfers and PGT, so you can see its practice pattern next to its outcomes.

Which measure answers which question

• “If I start a cycle at this clinic, how likely is a baby?” Look at per intended retrieval.
• “If I already have embryos and reach a transfer, how likely is it to work?” Per transfer is closer to your question, but it still depends on how the clinic selects embryos.
• “Across my whole treatment, what happened to people like me?” Look at the new patient measures, described below.

Why age matters more than almost anything else

Live-birth rates fall steadily with age. Nationally, the per-intended-retrieval rate goes from 49.7% under 35 to 7.7% over 40.^[2] A clinic that treats mostly younger patients will post higher overall numbers than one that treats many patients over 40, even if their care is equally good.

That is why you should always compare clinics within the same age group. The FertilityList Score goes a step further: it compares each clinic with the national rate for its own mix of patient ages. See the methodology.

The “new patient” measure

The CDC also reports, for first-time patients, the share who had a live birth after one, two, or all of their intended retrievals. Nationally, 60.8% of new patients under 35 had a live birth after all of their intended retrievals in the reporting period, compared with 53.4% after the first.^[2] This comes closest to the question “what happened to patients like me across their whole treatment?”

Small numbers and suppressed data

Some clinic figures carry symbols in the CDC data, and we show them the same way on clinic pages:

• Suppressed (*): the CDC hides counts of 1 to 4 to protect patient privacy.
• Fractions (†): when fewer than 20 cycles sit behind a rate, the CDC shows a fraction such as 3/12 instead of a percentage, because a percentage would be misleading.

When a clinic reports only a handful of cycles in an age group, one birth more or less can swing its rate by many points. FertilityList does not score clinics with fewer than 20 intended retrievals, and it pulls small samples toward the national average before comparing.

A checklist for comparing clinics

1. Use the same age group as yours.
2. Prefer per intended retrieval over per transfer.
3. Check the number of cycles behind the rate.
4. Look at more than one year. Every clinic page shows a three-year trend.
5. Remember the CDC’s caution: clinics serve different patient populations, so differences in rates do not by themselves prove differences in quality.^[1]

Ready to compare? Browse US fertility clinics by state. In 2022, US clinics reported 435,426 ART cycles to the CDC.