FSH, LH and Estradiol: What a Day-3 Hormone Panel Shows

It gives your doctor an estimate of how your ovaries are likely to respond to fertility drugs — roughly how many eggs a stimulated cycle might yield. It does not tell you whether you can get pregnant, and it cannot put a date on when you will stop being able to. The ASRM's 2020 committee opinion on ovarian reserve testing says this directly: these markers are useful for predicting oocyte yield after stimulation, but are poor predictors of reproductive potential independently from age.

That distinction is the whole point of the page. A high FSH is a reason to talk about timing and about which treatment makes sense. It is not a result that closes a door.

Here is what each of the three is actually for:

• FSH (follicle stimulating hormone) — the pituitary signal that recruits follicles. Measured early in the cycle, a higher value suggests the pituitary is pushing harder to get the same work done.
• Estradiol — the main estrogen the growing follicles make. Early in the cycle it should be low; if it is not, it changes how the FSH number should be read.
• LH (luteinizing hormone) — the signal that triggers ovulation mid-cycle. On day 3 it is background information; its useful moment is the mid-cycle surge, not the baseline.

Because a raised estradiol can pull a high FSH down into the normal range and hide it. Estradiol feeds back on the pituitary and suppresses FSH. If follicles have started growing unusually early, estradiol is already up on day 3, FSH gets suppressed, and the FSH figure reads reassuringly normal when the underlying picture is not.

Two IVF cohorts showed what happens when you look only at FSH. Evers 1998 followed 231 women through a first IVF attempt: among those whose day-3 FSH was normal, a high day-3 estradiol still marked a cancellation rate of 56% against 13% in those with a low estradiol, and a lower egg yield (9 versus 11). Smotrich 1995 reported the same pattern across 292 IVF cycles — and it held after excluding every woman with a raised FSH, so the estradiol signal was not simply standing in for FSH.

This is why a lab report showing a normal FSH and nothing else is an incomplete report, and why your doctor may want the pair repeated on the same blood draw.

FSH, LH and estradiol for this purpose are drawn in the early follicular phase — day 2 or day 3, counting day 1 as the first day of full bleeding, not spotting. Drawn later in the cycle the same numbers mean something completely different, because all three move across the month by design.

They also move between months. A single early-follicular panel is a snapshot, and doctors often repeat it in a later cycle before acting on a borderline result. If your report has one set of figures on it, you are looking at one cycle, not a settled fact about your ovaries.

One more limit worth knowing: the Endocrine Society's position statement on measuring estradiol (Rosner 2013) concluded that current assays are reasonably suited to diagnosing and managing infertility, but that imprecision and differences between laboratory methods remain problematic. Practically, that means a small change in an estradiol figure between two labs may be the method rather than you.

On day 3, LH mostly helps sort out where a problem sits — a low FSH and a low LH together point upwards, to the pituitary or hypothalamus, rather than to the ovary. LH earns its keep later in the cycle, when the surge that triggers ovulation is what home ovulation kits detect.

The LH-to-FSH ratio is widely quoted as a PCOS test, and it is worth being clear about its standing. The 2023 international evidence-based PCOS guideline does not include LH or the ratio among its diagnostic criteria. Those criteria are clinical or biochemical hyperandrogenism, ovulatory dysfunction, and polycystic ovarian morphology on ultrasound or — new in 2023 — an AMH level used in its place. LH is mentioned once in the whole recommendations document, and only about whether to add recombinant LH to FSH during IVF stimulation.

So a raised ratio on your report is not a PCOS diagnosis, and a normal one does not rule PCOS out.

Largely, yes, for that one job. Broer's 2013 individual patient data meta-analysis pooled 28 databases covering 5,705 women having IVF and compared how well each test predicted poor ovarian response. Age alone reached an area under the curve of 0.61. AMH reached 0.78 and antral follicle count 0.76, each a significant improvement on age. Combining AMH and follicle count did not improve on either alone.

The second half of that analysis is the part that rarely gets quoted. For predicting an ongoing pregnancy after IVF, age was the best single predictor at an AUC of 0.57 — weak — and none of the ovarian reserve tests added anything to it. The authors' own conclusion was that the clinical usefulness of these tests before IVF is limited to predicting ovarian response.

The 2023 PCOS guideline also draws a line around AMH: it recommends, strongly, that serum AMH should not be used as a single test to diagnose PCOS, and that AMH and ultrasound should not both be done, to limit overdiagnosis.

Setting out the limits plainly, because a hormone report is often read as more decisive than it is:

• It does not establish that you cannot conceive naturally. In Steiner's cohort of 750 women aged 30-44, a serum FSH above 10 mIU/mL was not associated with reduced fertility; neither was a low AMH nor a raised urinary FSH.
• It does not establish egg quality. These are quantity and responsiveness markers. No blood test in routine use reports on the quality of an individual egg.
• It does not establish when menopause will happen, and it is not a countdown.
• It does not establish a diagnosis on its own. FSH, LH and estradiol are read alongside your cycle history, an ultrasound, thyroid and prolactin, and a semen analysis for your partner.
• A single panel does not establish a trend. Values move between cycles, and between laboratory methods.

If the report has frightened you, that reaction is a reasonable response to being handed numbers without a frame for them, and not evidence that the numbers are bad.

The panel is one input. What typically follows depends on why it was ordered:

1. If cycles are irregular or absent, the search moves to why ovulation is not happening — thyroid function, prolactin, androgens and an ultrasound, rather than more gonadotropin testing.
2. If a borderline FSH or estradiol is the only abnormal finding, a repeat in a later cycle is common before any decision.
3. If IVF is being considered, AMH or an antral follicle count is the test that informs the drug protocol, and FSH becomes secondary.
4. If everything on the panel is normal and you have been trying for a year, normal hormones do not end the investigation — tubal patency and semen analysis are the usual next steps.

Sometimes the honest answer after a hormone panel is that the next step is testing or timing rather than treatment. A normal panel is not a reason to start IVF, and an abnormal one is not on its own a reason either.

• AMH explained — what your result means, and what it does not
• What is diminished ovarian reserve?
• Which hormone problem causes irregular periods
• What estrogen does across one menstrual cycle
• Understanding anovulation