How to Read CDC IVF Success Rates: Retrievals, Transfers and Cumulative Rates

By Berkcan K., editor · Published 2026-10-22 · Updated 2026-10-23

Every US fertility clinic is required to report each ART cycle it starts to the CDC, and the CDC publishes the outcomes clinic by clinic, one data year at a time. The latest final release is for cycles started in 2022, and the CDC notes that “pregnancy outcomes may have occurred in the selected year or the year after”, which is why the newest year is always about two years old. The report is honest but dense: the same clinic shows a dozen percentages, and the one a clinic quotes in its brochure is rarely the one that matters most to you. This guide walks the rows in the order the CDC prints them, using the measures and definitions on every clinic page of this site.

Three denominators, three very different rates

The CDC asks the same question, “what percentage resulted in a live-birth delivery?”, three times, each with a different base. The CDC glossary defines a live-birth delivery as “the delivery of one or more infants with at least one born alive.”

Per intended egg retrieval. The CDC counts a cycle started “with the plan to retrieve eggs (intended retrieval)”, whether or not the retrieval happened and whether or not any embryo was ever transferred. Cancelled stimulations, retrievals that produced no eggs, cycles stopped before transfer and cycles that banked everything for later all sit in the denominator. The rate is cumulative: the CDC’s own-egg success rates “include embryo transfers that occur within 1 year after an egg retrieval”, so a live birth from a frozen embryo transferred months later counts toward the retrieval it came from. In 2022 the national figure for patients under 35 using their own eggs was 49.7%.

Per actual egg retrieval. The CDC glossary defines egg retrieval as “a procedure to collect the eggs contained in the ovarian follicles.” This rate is cumulative in the same way (transfers within 1 year of the retrieval count) but leaves out cycles cancelled before the retrieval procedure, so it is never lower than the rate per intended retrieval. Nationally, 52.6% for patients under 35.

Per transfer. The glossary defines embryo transfer as “placement of embryos into a woman’s uterus through the cervix after IVF.” This rate counts each transfer on its own: the base is the number of transfers, and the outcome is the live birth from that one transfer. Cycles that never reached a transfer are not in the base, but a retrieval that led to two or three transfers puts two or three entries in it. That is why the per-transfer rate is not always the highest of the three: nationally in 2022 it was 49.1% for patients under 35, a little below the cumulative 49.7% per intended retrieval, in part because one retrieval with several transfers counts once as a success per retrieval but several times in the per-transfer base. In the older bands, where many cycles never reach a transfer, the per-transfer rate is usually higher: for patients over 40, 7.7% per intended retrieval against 25.4% per transfer. A per-transfer figure answers “once I have an embryo to transfer, what are the odds for that transfer?”, not “if I start a cycle, what are the odds?”

When a clinic’s marketing quotes one number, ask which of the three it is. The clinic pages here show all three side by side with the national figure under each.

Singleton rows

Each live-birth row has a singleton twin: the percentage that ended in “the delivery of a single infant born alive.” Twins and higher multiples carry more risk, so a clinic whose singleton rate is close to its overall rate is transferring one embryo at a time more often. A wide gap between the two rows is worth a question.

The cumulative rows: new patients

The second block covers patients with no prior ART at any clinic, followed through all the intended retrievals they had in the data year and any transfers of embryos from those retrievals, fresh or frozen. The CDC prints the live-birth percentage after one intended retrieval, after one or two, and after all of them, plus the average number of intended retrievals per new patient. This is the closest the report comes to “if I start treatment at this clinic, what is my chance of a baby?” Nationally in 2022, new patients under 35 reached a live birth 53.4% of the time after one intended retrieval and 60.8% after all retrievals in the year, with an average of 1.2 intended retrievals each.

Two cautions. The window is one calendar year, so a patient whose second retrieval fell in January of the next year is counted on the first. And a clinic whose new patients average 1.6 retrievals is not necessarily worse than one averaging 1.2; it may simply see older patients or harder diagnoses, which is exactly why the CDC splits everything by age.

The averages

Two rows are averages rather than percentages. “Average number of transfers per intended egg retrieval” shows how much embryo freezing a clinic does: a figure above 1 means retrievals typically yield more than one transfer. “Average number of intended egg retrievals per live-birth delivery” is the inverse of the first rate: a clinic at 2.0 needs two started cycles per baby in that age band. When a band had no live births at all the average cannot be calculated; the CDC file shows 0, and this site prints “No live births reported” rather than a 0 that would read as better than any real figure.

Donor eggs and donated embryos

Donor rates are not split by the patient’s age because the egg came from someone else and the donor’s age drives the outcome. Instead they are split by source: fresh embryos from fresh donor eggs, fresh embryos from frozen donor eggs, frozen embryos, and donated embryos. The rate is always per transfer. Nationally in 2022 the four were 52.0%, 45.2%, 45.1% and 42.2%.

Fractions, dashes and “no cycles reported”

The CDC’s data footnotes explain three conventions. “Fractions are used for rates with denominators less than 20”: a cell based on fewer than 20 retrievals or transfers is printed as a fraction such as 9/15. This site prints “9 of 15” and never converts it, because 9 of 15 is not 60% in any sense you can compare across clinics. An asterisk means “reported sample size of 1 through 4 have been suppressed”; values that could be used to work out a suppressed cell are suppressed too. This site prints a dash for those cells. Finally, where “no cycles meeting this criterion were reported”, the CDC file shows 0.0 with that footnote. That is an absence, not a 0% success rate, so this site prints “No cycles reported” and does not set the national figure beside it.

The denominators themselves are printed under each table: the number of intended retrievals, retrievals and transfers behind the rates. A 70% rate on 14 retrievals and a 48% rate on 400 are not the same kind of fact. Look at the count before the percentage.

What a success rate cannot tell you

The CDC’s guidance on interpreting the report is that “comparisons between clinics should be made with caution”, because “clinics may have different rates because they have different patient populations or ART treatment methods.” Small numbers make it worse: as the CDC puts it, if a clinic reports only one ART cycle in a category, its success rate there “would be either 0% or 100%.” Its success rates page adds that “average chances may not reflect actual success rates during ART treatment for an individual or couple,” and that chances “may vary according to age, type of infertility diagnosis, history of previous pregnancy” and the procedures used. Clinics also differ in who they accept: a program that takes patients other clinics turn away will show lower rates for that reason alone, and a program that steers patients with a poor prognosis toward donor eggs will show higher own-egg rates. The reasons-for-ART block on each clinic page (diagnoses as a share of cycles) is the best clue to the patient mix behind a rate.

Use the report to form questions for a consult: which rate is the clinic quoting, on how many cycles, for patients like you, and how does its cumulative new-patient rate compare with the national figure? That is what the tables on this site are built to answer; the decision itself belongs with you and a reproductive endocrinologist.