AMH vs. Antral Follicle Count for PCOS: What the 2023 Guideline Actually Changed

Which test does the 2023 guideline actually recommend — AMH or ultrasound?

The 2023 international guideline treats an AMH blood test and an ultrasound follicle count as two routes to the same finding, not a package deal: its own text states that “either serum AMH or ultrasound may be used to define” polycystic ovarian morphology (PCOM), “however, both tests should not be performed to limit over-diagnosis,” under the 2023 International Evidence-Based Guideline. That is a genuine change from how the 2003 Rotterdam criteria worked, when ultrasound was the only way to establish that one feature. The guideline’s own discussion section calls this “a rapidly evolving area” since 2018 that is “now strong enough” to support the new recommendation, and says it “will significantly change practice and offers women a low cost, convenient option, without evidence of overdiagnosis.”

That choice only comes up partway through the full PCOS diagnostic work-up, for a subset of people being evaluated. Two of three Rotterdam features are still required — irregular ovulation, androgen excess, and polycystic ovarian morphology — and where irregular cycles and androgen excess are already both present, the guideline says outright that an AMH level “is not necessary for PCOS diagnosis” at all, because two of the three boxes are already checked. The AMH-versus-ultrasound decision applies to the narrower group where polycystic ovarian morphology still needs to be established — most often someone with irregular cycles but no clear androgen excess, or the reverse.

Note: in May 2026, PCOS was renamed polyendocrine metabolic ovarian syndrome, or PMOS, by a global consensus of more than 50 organisations. Nothing about which test a clinician chooses changed with the name; this article uses PCOS because that is still the term most readers search.

What’s the practical difference between a blood draw and a follicle-count ultrasound?

A blood draw for AMH needs a needle, a lab that runs the assay, and a few minutes — a transvaginal ultrasound needs a probe running at 8 MHz or higher and someone trained to operate it and count follicles correctly. That gap in equipment and personnel is exactly what the guideline is pointing at when it calls AMH “a low cost, convenient option”: one test can be drawn at almost any clinic or lab-draw site, while the other requires ultrasound equipment and a sonographer, radiologist, or gynecologist to perform and interpret it.

Why does a blood test now count the same as an ultrasound in adults?

A 2024 meta-analysis pooling 68 adult studies found AMH identifies PCOS with 79% sensitivity (95% CI 76–82%) and 87% specificity (95% CI 84–89%), which is the evidence base the 2023 guideline leans on to accept AMH as an alternative to ultrasound rather than a downgrade from it. The two measurements track each other because they’re reading the same biology from different angles: AMH is produced by the granulosa cells lining small antral follicles, and PCOS ovaries carry roughly two to three times the normal number of those follicles, a difference documented directly by comparing 59 people with PCOS against 45 controls. An ultrasound counts the follicles by sight; a blood test measures the hormone those same follicles are producing.

It’s worth knowing that the ultrasound side of this comparison isn’t a fixed gold standard either. A 2013 study of 168 women — 98 with PCOS by NIH criteria and 70 controls — found that newer, higher-resolution transducers required raising the follicle-count threshold to keep the test accurate, because better equipment could see follicles older machines missed. The 2023 guideline settled on 20 or more follicles per ovary on an 8 MHz-or-higher probe, but it says that threshold “should be revised regularly with advancing ultrasound technology” — the same kind of moving target that gets raised as an objection to AMH.

Why is there no single AMH number, and does that make ultrasound more reliable?

There is no international AMH cut-off because the guideline’s own working group instructs laboratories to “use population- and assay-specific cut-offs” rather than one global number, under the 2023 guideline’s practice points — different AMH assay platforms measure the same blood sample differently enough that a single universal threshold would misclassify people depending on which lab ran the test. That is the reason your own lab report carries its own reference range rather than a number you can look up online. AMH reference ranges and what they mean are covered on their own, separately from the diagnosis question here.

Ultrasound is not immune to the same problem, just a different version of it. The threshold only holds for a transducer of 8 MHz or higher, and a scan on older or lower-resolution equipment can under-count follicles enough to miss the finding entirely. Neither test hands a clinician one context-free number — an AMH result is only meaningful against your lab’s own cut-off, and a follicle count is only meaningful once you know what equipment produced it. The full ultrasound follicle-count thresholds are laid out elsewhere; the point that matters for choosing between the two tests is that “more visual” does not mean “more standardized.”

Why doesn’t any of this apply if you’re within 8 years of your first period?

Neither AMH nor ultrasound is used to diagnose PCOS within eight years of menarche, and the guideline’s reasoning is specific: performance on both tests drops enough in adolescence that it stops being useful for telling PCOS apart from normal puberty. The same 2024 meta-analysis found AMH’s accuracy fell to 66% sensitivity and 78% specificity in adolescent studies, well below adult performance, and the guideline states plainly that “there are no definitive criteria to define polycystic ovary morphology on ultrasound in adolescents; hence, it is not recommended” at that age either. Its discussion section is more direct still, describing both ultrasound and AMH as “not recommended due to poor specificity” in adolescents specifically.

The biology explains why. Multi-follicular ovaries and naturally elevated AMH are ordinary features of a maturing reproductive system in the years following a first period, not a sign of disease — which means a scan or a blood draw that would flag PCOS in an adult can be a normal finding in a teenager. For adolescents, the guideline instead requires both androgen excess and ovulatory dysfunction together, with no substitute for either one, and suggests an “at risk” label with re-evaluation around 8 years post-menarche where only one feature is present. If you’re a teenager reading this because someone mentioned AMH or a follicle count, the honest answer is that this entire blood-test-versus-ultrasound decision does not apply to you yet — your diagnostic path runs through symptoms and cycle history, not either of these tests.

What neither test can tell you

Neither an AMH result nor a follicle count diagnoses PCOS by itself, in anyone of any age. Each one substitutes for exactly one of three Rotterdam features — polycystic ovarian morphology — and still has to be paired with either irregular ovulation or androgen excess to complete the two-of-three rule. An elevated AMH or a follicle count above threshold, on its own, with regular cycles and no androgen excess, does not meet the definition. This article reports what the criteria are; it does not — and cannot — tell you whether your own number crosses into “you have PCOS.” That judgment depends on your full history, your exam findings, and ruling out other causes, which is a clinician’s job working from the whole picture, not a threshold a reader can check against alone.

It’s also worth being precise about what “antral follicle count” means in each setting, because the same phrase gets used two different ways. The diagnostic count discussed here looks at one ovary at a time and asks whether it meets the polycystic-ovarian-morphology threshold. A fertility clinic’s AFC report is a different, bilateral number built to plan how you might respond to ovulation-induction or IVF medication — antral follicle count in fertility treatment is its own measurement, not a diagnostic test at all. Confusing the two is an easy, common mistake, and neither version predicts egg quality or how soon you might conceive.

Your next step

If you’re being worked up for PCOS and polycystic ovarian morphology is still an open question, ask your clinician directly which of the two tests they plan to use, and why — not both, per the guideline, so if both get ordered it’s worth asking. If it’s AMH, ask which assay the lab used and what cut-off it applies, since that number is lab-specific. If it’s ultrasound, ask what probe frequency was used and which cycle day you were scanned on. And if you’re within 8 years of your first period, this choice isn’t the one your clinician should be making at all — ask instead which of the two mandatory adolescent features, androgen excess or ovulatory dysfunction, is still being assessed.