GnRH Antagonist vs Agonist Protocol — Which IVF Protocol, and Why (2026)

For most people, the antagonist protocol. In the main Cochrane review the chance of a baby was similar with either, but the antagonist protocol carried a lower risk of ovarian hyperstimulation syndrome (OHSS) and needs no weeks of suppression before stimulation starts. Europe's specialist society, ESHRE, recommends it for the general IVF population in its 2025 guideline, and the UK's NICE adopted those recommendations in 2026.

That is a recommendation, not a verdict on the long agonist protocol. One careful review found slightly fewer ongoing pregnancies with the antagonist in some patients, and there are people for whom the long protocol is a reasonable choice. What you are owed is a reason for the protocol you are on.

It is easy to feel that the protocol is the one decision that will make or break the cycle. On the evidence it rarely is — which should make it easier, not harder, to ask about.

Daily hormone injections (gonadotropins) make several follicles grow at once. Left alone, the brain would release a surge of luteinising hormone and the eggs would ovulate before they could be collected. The "protocol" is mostly the method used to stop that surge. The overall sequence is set out on our IVF treatment page.

• Antagonist ("short") protocol. An injection that blocks the surge directly and works within hours, so it is added part-way through stimulation — either on a fixed day (day 6 or 7) or once the leading follicle reaches about 14–15 mm. There is no suppression phase beforehand.
• Long agonist protocol. A different drug that first stimulates and then switches off the pituitary. It is started before stimulation and continued through it; one trial in a 2025 Cochrane review compared giving it for two weeks versus three before stimulation began.
• Progestin-primed stimulation (PPOS). Progesterone-type tablets taken through stimulation instead of either injection. Covered further down.

The stimulation drugs and doses can be the same in either protocol. So the protocol does not decide how good the eggs are; it decides how ovulation is held back, and with it the OHSS risk and which trigger injection can be used at the end.

Not clearly. The 2016 Cochrane review found no conclusive difference in live birth, and a 2025 Cochrane network analysis of 338 trials reached the same place for women with a normal expected response: little to no difference in live birth or ongoing pregnancy (risk ratio 0.95, 95% CI 0.84 to 1.07; 8 trials, 2,817 women; moderate certainty).

The honest complication is a 2017 review by Lambalk and colleagues that grouped trials by patient type. In general IVF patients, ongoing pregnancy was slightly lower with the antagonist (risk ratio 0.89, 95% CI 0.82 to 0.96). The authors put it as one fewer ongoing pregnancy for every 28 women treated, against one case of OHSS prevented for every 40. The gap was concentrated in the flexible-start schedule combined with a contraceptive pill beforehand; with a fixed-start schedule there was no significant difference.

The largest single trial, 1,050 women under 40 in their first cycle, found live birth of 22.8% per woman randomised on the antagonist and 23.8% on the agonist — a difference that could easily be chance.

You may collect slightly fewer eggs on the antagonist: the 2025 Cochrane analysis estimated about 0.75 fewer for normal responders, on low-certainty evidence. More eggs can mean more embryos to freeze, which matters if you hope for more than one child. But in these trials fewer eggs did not translate into fewer babies, so egg count alone is a poor reason to choose.

The antagonist protocol. OHSS, an over-response of the ovaries, is the main serious complication of stimulation. Pooling 36 randomised trials of 7,944 women, Cochrane estimated that a risk of about 11% on the long agonist falls to between 6% and 9% on the antagonist. In the trial of 1,050 women in their first cycle, severe OHSS affected 5.1% of women versus 8.9%, and 1.7% versus 3.6% were admitted to hospital with it. Symptoms and prevention are covered in our guide to IVF side effects and OHSS.

A large part of the safety advantage is the trigger. Because the pituitary has not been switched off in advance, an antagonist cycle can end with a GnRH agonist trigger instead of hCG. Cochrane found this cuts OHSS sharply (odds ratio 0.15) — but it also lowered live birth when embryos were transferred fresh in the same cycle (odds ratio 0.47: a 31% chance with hCG became 12–24%). That is why ESHRE recommends an agonist trigger for women at risk of OHSS together with freezing all embryos and transferring them in a later cycle.

If you have been told you are at high risk — usually because of a high AMH or antral follicle count, as in many women with PCOS — ask what the trigger and transfer plan is, not just the protocol. Our OHSS risk-factor tool lists what clinics usually weigh.

ESHRE sorts patients by how strongly their ovaries are expected to respond, and it recommends antral follicle count or AMH — not age or FSH — to predict that.

• Expected high response, including many women with PCOS. ESHRE strongly recommends the antagonist protocol, and probably a reduced starting dose (100 to under 150 IU). In Lambalk's review, women with PCOS had no difference in ongoing pregnancy (RR 0.97) and less OHSS (RR 0.53) on the antagonist. More in our PCOS article.
• Expected normal response. The antagonist protocol, at a conventional dose, which ESHRE now defines as 150–225 IU; it does not recommend going higher or lower routinely.
• Expected low response. No difference in ongoing pregnancy between the protocols in Lambalk's review (RR 0.87, six trials). ESHRE does not recommend doses above 300 IU. See our diminished ovarian reserve article.
• Egg donors and egg freezing. ESHRE strongly recommends an agonist trigger for egg donors on antagonist or progestin protocols, and for elective egg freezing it probably recommends antagonist or progestin protocols over agonist ones.

The long agonist protocol keeps a place. ESHRE says that if an agonist is used, the long version is recommended over the short or ultrashort ones. The point is that the choice should come with a reason specific to you, such as how you responded in an earlier cycle.

PPOS. Tablets are easier than injections and could be cheaper, which is its appeal. A 2023 Cochrane review of 14 trials (2,643 women having IVF, plus egg donors) found there may be little or no difference in cancelled egg collections compared with antagonists — though even that was uncertain — and the evidence on live birth was too uncertain to draw conclusions. ESHRE 2025 rates it "probably equally recommended" only where all embryos are to be frozen and transferred in a later cycle, on very-low-quality evidence. Ask whether a freeze-all is the plan, and what the frozen transfer will add in time and cost.

Mild stimulation. ESHRE dropped the term from its 2025 guideline: studies used it for very different doses, some no lower than conventional stimulation, so they could not be compared. If a clinic offers "mild IVF", ask for the actual daily dose.

Natural or modified natural cycle. ESHRE says a modified natural cycle is probably not routinely recommended over conventional stimulation for low responders, while accepting that for women with very low reserve a clinician could choose it. Separately, the 2025 Cochrane analysis found that stimulation without any pituitary suppression may lower live birth or ongoing pregnancy compared with an antagonist protocol (RR 0.71; low certainty).

India's ART Act requires clinics to explain the advantages, disadvantages, cost, side effects and risks of each procedure (section 21(c)), so these are ordinary questions:

1. Which protocol am I on, and what about my tests or history led to it?
2. Am I expected to be a high, normal or low responder, and on what result — AMH or antral follicle count?
3. What is my OHSS risk, and what is the plan if I over-respond — which trigger, and fresh or frozen transfer?
4. What is my starting dose, and why that dose?
5. What will the drugs cost, itemised?

The last question is worth asking because drugs are the biggest cost. An ICMR study of 148 couples at five tertiary hospitals found an average out-of-pocket spend per IVF cycle of ₹2,30,668 in private hospitals and ₹1,10,104 in public ones; across the 148 couples, drugs were 55% of total spending. We found no Indian data comparing the cost of one protocol with another.

And if you are earlier in the process, the right next step may not be IVF at all. Sometimes the answer is testing, timing or a simpler treatment. Our guide to reading IVF success rates helps with the numbers you are likely to be quoted.

A disclosure: IVF Pulse runs fertility clinics, which also choose protocols for their patients. The evidence above applies to them as much as to anyone.