AMH in PCOS: What the Number Means and Why It Is Not a Diagnosis

What AMH level indicates PCOS?

There is no single AMH number that confirms PCOS on its own — it is one input into the full diagnostic work-up, not a stand-alone test. A 2024 meta-analysis pooling 68 adult studies found that AMH testing distinguishes PCOS from non-PCOS with 79% sensitivity (95% CI 76–82%) and 87% specificity (95% CI 84–89%), which is accurate enough that the 2023 guideline now accepts an elevated AMH as a substitute for a pelvic ultrasound when confirming polycystic ovarian morphology in adults. That same review explicitly declined to set one universal cut-off, because AMH assays vary enough between laboratories that a single global threshold would misclassify too many people. In practice this means your lab’s own cut-off — printed on your report — is the number that applies to your result, the same way a testosterone or DHEA-S reference range is lab-specific rather than universal.

Note: in May 2026, PCOS was renamed polyendocrine metabolic ovarian syndrome, or PMOS, by a global consensus of more than 50 organisations. AMH’s role in the diagnostic algorithm did not change with the name — this article uses PCOS because that is still the term most readers search.

Why isn’t AMH a PCOS diagnostic criterion on its own?

Because the same 2024 meta-analysis that supported using AMH as an ultrasound substitute also concluded that AMH levels alone are insufficient for PCOS diagnosis and are nonspecific for polycystic ovarian morphology in adolescents — in that age group, sensitivity and specificity dropped to 66% and 78%, well below the reliability seen in adults, which is why the 2023 guideline does not use AMH for diagnosis under 20 at all. Even in adults, AMH only ever substitutes for one leg of the Rotterdam criteria — the ultrasound-equivalent evidence of polycystic ovaries. It still needs to be paired with either irregular ovulation or hyperandrogenism to complete the two-of-three rule; a high AMH by itself, with regular cycles and normal androgens, does not meet the definition.

What are normal AMH levels by age?

AMH rises through the teens and twenties, then declines gradually with age in everyone, becoming undetectable roughly five years before menopause as the pool of ovarian follicles is depleted, according to a 2014 review of the ovarian-reserve literature. In PCOS, that same decline runs from a higher starting follicle pool, which is part of why cycles often regularize in the 40s — see PCOS in perimenopause for what changes at that stage and what does not. Labs report AMH in either nanograms per millilitre (common in the US) or picomoles per litre (common in the UK, EU, and Australia); the conversion is 1 ng/mL ≈ 7.14 pmol/L.

These are commonly published ranges from general laboratory sources, not a single validated study, and they are not PCOS-specific — a person with PCOS at any age typically sits well above their age-matched band, sometimes several multiples higher, which is the entire basis for using AMH as a marker in the first place. Every lab sets and prints its own cut-off; compare your result to that number, not to this table.

What does antral follicle count have to do with PCOS?

AMH is produced by the granulosa cells lining small antral follicles, and PCOS is associated with roughly a 2- to 3-fold increase in the number of those small growing follicles compared with ovaries without PCOS — which is mechanically why AMH runs high in PCOS and why the two measurements track each other closely on ultrasound. That original 2003 study, comparing 59 people with PCOS to 45 controls, found AMH correlated tightly with directly counted follicle number, establishing AMH as a blood-based proxy for a finding that would otherwise require a transvaginal scan to see. A high AMH in PCOS is, mechanically, a report of follicle quantity — specifically, an excess of small follicles arrested early in development — not a report of egg quality, egg count available for retrieval, or how close you are to menopause.

Does a high AMH in PCOS mean better fertility — or does a normal AMH mean you’re fine?

Neither, and this is the point most likely to get garbled online. A 2017 JAMA cohort study of 750 women aged 30 to 44 trying to conceive — a general population, not a PCOS-specific one — found that women with low AMH (below 0.7 ng/mL) had essentially the same probability of conceiving within 12 cycles (84%) as women with normal AMH (75%), concluding that AMH should not be used to assess natural fertility. If a marker purpose-built to flag low ovarian reserve does not reliably predict who conceives and who doesn’t in the general population, an elevated AMH in PCOS — which reflects an unusually large pool of small, often-arrested follicles rather than a reassuringly large pool of ready-to-ovulate ones — cannot be read as a fertility bonus either. What actually predicts whether ovulation is happening is covered separately in how PCOS affects egg quantity and quality, which AMH does not measure. Whether metformin, the pill, inositol or weight loss can actually lower that number is a related but separate question: trial evidence on AMH and PCOS treatment walks through what the data actually shows.

Who this doesn’t help

An AMH result will not tell you whether you are ovulating this month, whether you will conceive this year, or how many eggs a retrieval would produce — none of those questions are what AMH measures. It also will not settle a diagnosis on its own: someone with an elevated AMH, regular cycles, and normal androgens does not meet the Rotterdam criteria on that basis alone, since AMH only substitutes for the ultrasound leg. And for anyone under 20, AMH is not validated for diagnosis at all, elevated or not — a normal AMH in an adolescent with clear symptoms should not be read as reassurance, and a high one should not be read as confirmation, per the 2023 guideline’s own limits on the test. If your labs come back with a normal AMH alongside other findings, normal results don’t automatically rule PCOS out either — it simply means that particular leg of the criteria has to be assessed another way.