The number every clinic quotes, and how to read it.
Success rates are the most quoted and least explained figure in fertility treatment. This page gives you the age-banded ranges, shows exactly how such a number is built - and where it can be inflated - and lets you work out a realistic range for yourself.
Not a One World Fertility outcome claim. Registry-typical bands, Indian & international ART reporting, reviewed January 2026. Published so you can calibrate any clinic’s figures - including ours.
-
47%IVF, age 30-34 Live birth per started cycle, midpoint of the published band.
-
85%After three cycles Chance of at least one live birth, same age band, cycles treated as independent.
-
54%Donor eggs, 35-37 Barely moves with your age, because the eggs are not yours.
-
6Age bands reported Every figure on this page is broken out by age. None is a clinic-wide average.
What is a realistic IVF success rate?
Measured the honest way - live birth per started cycle - own-egg IVF runs at roughly 52-58% under 30, 44-50% at 30-34, 33-39% at 35-37, 21-27% at 38-40 and 4-7% at 43 and over. Donor-egg cycles stay near 46-54% at every age, because the eggs are not the patient’s. Rates quoted per transfer are always higher than rates quoted per started cycle and the two must never be compared.
- Outcome counted
- Live birth - not positive beta, not clinical pregnancy
- Denominator
- Cycles started, including those cancelled before retrieval
- Age bands
- 6, reported separately - never one clinic-wide figure
- Pathways
- 5 - IVF, ICSI, frozen transfer, IUI, donor egg
- Embryos per transfer
- One, in almost every case
- Source
- Registry-typical bands, Indian & international ART reporting
- Reviewed
- January 2026
Live birth per started cycle, by age
Each bar is a range, not a point. Pick a pathway; add a second to compare.
Every figure on this page, in one table
| Pathway | Under 30 | 30 to 34 | 35 to 37 | 38 to 40 | 41 to 42 | 43 & over |
|---|---|---|---|---|---|---|
| IVF | 52-58% | 44-50% | 33-39% | 21-27% | 10-15% | 4-7% |
| Frozen transfer | 50-57% | 43-50% | 34-40% | 23-29% | 12-17% | 5-8% |
| Donor egg IVF | 50-58% | 50-58% | 50-58% | 50-58% | 48-56% | 46-54% |
| IUI | 12-18% | 10-15% | 7-11% | 4-7% | 2-4% | 1-2% |
| IVF with ICSI | Reported inside the IVF figures. ICSI is a fertilisation technique, not a separate outcome curve - quoting it as its own success rate would be counting the same cycles twice. | |||||
Why one number is never the answer
Four ways a success rate can be true and useless at the same time.
- Per started cycle 33% 100 cycles begun, 33 babies
- Per egg retrieval 41% Cancelled cycles quietly removed
- Per embryo transfer 52% Everyone who reached no transfer removed too
- “Pregnancy rate” per transfer 68% A positive test, not a baby - the banner number
Illustrative arithmetic, not a real clinic. Every one of these four is defensible in a footnote. Only the first answers “if I start, what happens?”
-
The denominator
Per started cycle, not per transfer
Divide the births by the transfers and you have quietly deleted every cycle that was cancelled, every retrieval that produced no usable embryo and every patient who never reached the table. Divide by the cycles that were started and the number gets smaller, harder and honest - because that is the decision you are actually making.
-
The denominator
Who was allowed into the count
A clinic that declines difficult cases, or quietly moves them to donor eggs before counting, will publish a better headline than the clinic that treats them. Nothing in the arithmetic is false. It is selection, and it is invisible unless outcomes are reported by age band and by cause.
-
The definition
Pregnancy is not a baby
Positive beta hCG, clinical pregnancy on a scan, ongoing pregnancy and live birth are four different numbers, and they fall in that order. A rate quoted without saying which one it is will almost always be the largest one available.
-
The arithmetic
One cycle or all of them
Cumulative figures across three cycles look far better than a single cycle, and they are legitimate - as long as it says so. Read carefully, "up to 70%" is often three cycles, best age band, per transfer, all at once.
How a success rate is actually built
Follow 100 started cycles through to delivery. Every step is where a percentage point is lost - or quietly removed from the denominator.
-
100of 100 Cycles started
One hundred people begin stimulation. This is the denominator that matters, and the one most often replaced.
-
92of 100 Reach egg retrieval
A small number are cancelled - a poor response, an over-response risking OHSS, or ovulation before the retrieval could happen.
-
86of 100 Have a fertilised egg
Eggs are collected, but not all are mature and not all mature eggs fertilise. Total failed fertilisation is uncommon and never expected.
-
71of 100 Reach an embryo transfer
Some embryos arrest before day 5. That is the laboratory doing its job - finding the embryos with a real chance rather than transferring one that was never going to implant.
-
47of 100 Take a baby home
Counted across the fresh transfer and the frozen transfers from that same egg collection. This is why a per-started-cycle figure can sit above a per-transfer one, and why the two must never be compared directly.
A worked illustration, pinned to the chart above. The final step is the midpoint of the published 30 to 34 IVF range (44-50%), so this infographic can never contradict the data. The attrition at each earlier step is typical rather than measured, and every step shifts with age.
Now delete the first two steps. The same 100 cycles become 71 transfers and 47 babies - a rate of 66% “per transfer”. Nothing was falsified. The denominator moved, and the headline gained 19 points.
Where the outcome is actually decided
Nine stages between the first blood test and a delivery. Most of the number is set before you inject anything.
-
BeforeDay −60
The workup
AMH, antral follicle count, a semen analysis and an assessment of the uterine cavity. More of your final number is decided here than at any later point, and none of it is under a clinic’s control - only its honesty about what it is seeing.
Sets the band you start in -
Days 1-12Stimulation
How the ovaries answer
Scans and estradiol every few days. The protocol is adjusted to the response rather than the plan. A cycle cancelled here is counted as a failure per started cycle - which is precisely why that denominator is the harder one.
Protocol adapts to you -
Day 14Retrieval
Mature eggs, not eggs
Only mature eggs can be fertilised, so the number collected and the number usable are rarely the same. This is the first hard number in the cycle and you are told it the same day.
First honest count -
Day 15Fertilisation
Two pronuclei
Sixteen to eighteen hours after ICSI or insemination, each egg is checked. Two pronuclei is right; none means fertilisation failed and three means it went wrong. This is the second call you get.
Second hard number -
Day 5Blastocyst
How many made it
Ten eggs can become two blastocysts, and that is normal rather than a disaster. Embryos that arrest in the laboratory would overwhelmingly have failed in the uterus, where you would have paid for it in weeks of waiting instead of days.
The real predictor -
TransferDay 5 or FET
Embryo meets endometrium
One embryo, in almost every case. Two does not double your chance of a baby; it multiplies the chance of a twin pregnancy, which is the single largest avoidable risk in fertility treatment.
Single transfer by default -
+12 daysBeta hCG
The first number that is not a rate
A blood test, not a home kit. A positive here is a biochemical pregnancy and nothing more - it is the point where the four different definitions of “success” begin to separate.
Where definitions diverge -
Week 6-7Viability scan
A heartbeat
Clinical pregnancy is confirmed on ultrasound. Most clinic “success rates” stop counting here, because this is where the number is at its most flattering.
Where most rates stop -
DeliveryWeek 38-40
Live birth
The only outcome that answers the question you actually asked. Every figure on this page is measured here, which is why they are lower than the numbers you will see elsewhere.
The only figure we publish
Estimate your own range
Seven selections. The readout updates live and always returns a range, because a single number would be a lie about precision.
This is the quick version, and it runs on the same model. The full IVF Success Rate Estimator walks you through the same factors one at a time and explains what each one is doing to your number.
chance of a live birth from one started cycle
- 1 cycle 44-50%
- 2 cycles 68.6-75%
- 3 cycles 82.4-87.5%
- Nothing yet - every selection is at its neutral setting.±0%
This is not a prediction, and it is not our result. It applies published, indicative modifiers to registry-typical age bands. It cannot see your AMH value, your antral follicle count, your embryo grades or your uterine cavity - and those decide the outcome. Bring it to a consultation as a question, not as an answer.
Seven questions for any clinic’s number
Including ours. If a clinic cannot answer the left column, the figure on its banner is marketing.
- What is the denominator? Cycles started, including the ones cancelled before retrieval. Embryo transfers - which silently removes every cycle that never got that far.
- Which outcome is being counted? Live birth. A baby, delivered. “Pregnancy rate” or “positive beta”, both of which are larger and neither of which is a baby.
- Is it broken out by age? Six bands, reported separately, on this page. One clinic-wide figure, which mostly measures who the clinic agreed to treat.
- One cycle or several? Per cycle by default; cumulative figures are labelled as cumulative. “Up to 70%” - three cycles, best age band, per transfer, all at once, none of it stated.
- Are donor cycles counted in? Reported as their own pathway, never blended into the own-egg figure. Mixed into a single headline, which lifts the older age bands substantially.
- How many embryos per transfer? One, in almost every case, and the rate is quoted on that basis. Multiple embryos transferred - a better headline bought with a twin-pregnancy risk you carry.
- Where do the numbers come from? Named source, stated period, stated method - as at the top of this page. No source, no period, no denominator. A number on a banner.
People also ask
The questions that come up once the percentages are on the table.
Ask us for your number, not the average.
Bring your AMH, your antral follicle count and any previous cycle reports. You will leave with a range calculated from your own results, in writing, along with what would have to change to move it.