What actually decides whether IVF works
The biggest factor in whether IVF works is your age, by a wide margin, more than diagnosis, more than clinic choice, more than most of what you can control. This page explains what actually drives the odds using the largest published model of its kind, 91,000 UK women tracked all the way to live birth, shows real numbers by age, covers why looking across more than one treatment cycle changes the picture, and explains why a low AMH is not a reason to be pushed toward donor eggs.
Written by Fertility Decoded. Reviewed by D R, MS (OBG), FRM. Review date: 29 August 2026.
The single biggest factor in whether IVF works is your age, by a wide margin. Most of what you can influence matters far less than you have probably been told, and a clinic's advertised success rate says more about how it counts than about your own chances.
Get your own estimate, built on the same published model this page explains, computed from your own details rather than one clinic's headline number.
What actually drives your odds
In the largest published model built for this, a woman's age accounts for about 85% of everything the model can explain. Duration of infertility, diagnosis, and whether she has been pregnant before all matter, but they are small next to age.
Clinic choice is real, but it is second-order. In the source study, clinics at the low end of the range had roughly 0.6 times the average odds, and those at the high end roughly 1.65 times, after adjusting for patient characteristics. That is a genuine difference worth caring about. It is not the difference between a 20% clinic and an 80% clinic. Most of your outcome is set before you walk in the door, and clinic choice adjusts it rather than defines it.
Not every diagnosis carries the same outlook either, and the ordering surprises people. Of the four causes this kind of model measures, ovulation problems carry the most favourable outlook and tubal factor the least, with male factor and unexplained infertility in between. A woman told she has PCOS usually hears it as bad news. Learning that, among these four causes, it carries a relatively better IVF outlook is worth knowing. Keep this in proportion: it is a few points of difference, not a transformation. The 2016 study behind this model found couples with unexplained infertility or anovulation had slightly higher chances of success.
One cycle is the wrong unit
A complete cycle is one egg collection plus every fresh and frozen transfer that results from it. Clinics can present this figure in ways that flatter their numbers, which questions to ask your clinic and red flags to watch for both cover in detail. What matters here is what happens when you look across more than one complete cycle rather than just one.
For a 34-year-old, one complete cycle gives a 36% chance of a live birth. Across three complete cycles, that rises to 68%. These describe different plans, and the second, the cumulative figure, is the honest one to decide against: it means budgeting and preparing for a course of treatment rather than a single attempt, financially and emotionally. Trying again after a cycle that did not work goes through what that actually involves.
What the numbers actually look like
Live birth, own eggs, average patient profile, UK data 2010 to 2016:
| One complete cycle | Across three complete cycles | |
|---|---|---|
| 30 | 42% | 75% |
| 32 | 39% | 72% |
| 34 | 36% | 68% |
| 36 | 32% | 62% |
| 38 | 24% | 50% |
| 40 | 14% | 31% |
| 42 | 7% | 16% |
| 44 | 3% | 8% |
These are group averages for an average patient profile at each age, not a prediction about you. Your own duration of infertility, diagnosis, and pregnancy history shift the number in either direction. Get your own figure from the tool using your actual details.
Your AMH is not your destiny
AMH predicts how many eggs are likely to be collected. It is a poor predictor of whether a cycle ends in a baby. The strongest evidence for this is structural: the largest externally validated model of cumulative live birth contains no measure of ovarian reserve at all, no AMH and no antral follicle count, and it still predicts live birth to a useful standard without them.
A low AMH on its own is not a reason to be pushed toward donor eggs. If a clinic raises donor eggs on the strength of an AMH result alone, the questions worth asking first apply here as much as anywhere.
Why we cannot show you an Indian number
The statutory National ART and Surrogacy Registry publishes clinic registration, not treatment outcomes. The last multi-centre Indian dataset with real reach, from NARI and ISAR using 2007 to 2009 data and published in 2013, reports pregnancy rate rather than live birth, and its own authors say age stratification was impossible because clinics did not submit age data. One Indian cohort clears a real quality bar: a single Bangalore hospital's series, published in the Journal of Human Reproductive Sciences in 2022, covering 1,989 cycles. It is one hospital's own results, not a national figure, and we label it as such wherever it is cited.
The most authoritative Indian source available is missing the exact things red flags to watch for already tells you to demand from a clinic: a live-birth rate, broken down by age. Until that changes, the numbers on this page and in the tool describe British patients.
What to do with this
Questions to ask your clinic covers what to raise before you commit. Red flags covers the warning signs that a clinic is not being straight with its own numbers. Your own chances gives you a figure built from your details rather than a headline claim. And if a cycle does not work, trying again walks through the cumulative picture and what is actually worth changing.
- It depends entirely on what it is a rate of. A 50% live-birth rate per complete cycle for a woman in her early thirties using her own eggs would be unusually high. The same number as a pregnancy rate, blended across ages, or counted per embryo transfer instead of per cycle, means much less. Ask what exactly is being measured before judging the number.
- Not reliably. Clinic choice does shift outcomes somewhat, but age and diagnosis matter far more. A high price is not evidence of a better lab, and it is not something this page or the tool can verify for any specific clinic.
- There is no single answer, but the cumulative figures on this page show why one cycle rarely tells the whole story. Chances rise substantially across a second and third complete cycle for most ages. Trying again after a cycle that did not work covers what is actually worth changing in between.
- No. AMH predicts how many eggs are likely to be collected, not whether a cycle succeeds. The model behind this page's numbers does not use AMH at all and still predicts live birth to a useful standard.
- We do not know, and neither does anyone else with confident, national data. India has no published equivalent to the UK study behind these figures. Treat the numbers here as the best published estimate available, not a claim about Indian clinics specifically.
- One egg collection plus every fresh and frozen embryo transfer that results from it. It is a more honest unit to plan around than a single transfer, because it reflects what one round of stimulation and retrieval can actually give you a chance at.
Is a 50% success rate good?
Does a more expensive clinic mean better odds?
How many cycles do most people need?
Does my AMH mean IVF will not work?
Are Indian success rates different from what is shown here?
What is a complete cycle?
Next: get your own estimate from the tool, or read how we calculate it.
Sources and further reading
- McLernon and colleagues: Cumulative IVF model (BMJ, 2016). Original model and analysis of age and clinic effects.
- Ratna and colleagues: External validation (Human Reproduction, 2023). Updated UK model; site calculations are explained in our methodology page.
- ASRM: Ovarian reserve testing (2020). AMH prediction limitations.
- Malhotra and colleagues: Indian ART data, 2007–2009 (2013). Historical multicentre Indian reporting, not current national live-birth rates.
- Indian single-centre cumulative live-birth study (2022), PubMed abstract. Single-hospital cohort of 1,989 women, not national data.
Source links checked on 7 September 2026. This source check is separate from medical review.
Related
Last updated .