Antral Follicle Count in PCOS: Why the Threshold Isn't Fixed

What is antral follicle count, and how is it different from the PCOS diagnostic count?

Antral follicle count (AFC), as fertility clinics use the term, reports one number: the total of small, hormone-responsive follicles visible across both ovaries on an early-cycle transvaginal scan. That is a structurally different measurement from the count the PCOS ultrasound criteria actually use for diagnosis, which looks at one ovary at a time and requires 20 or more follicles in at least that single ovary under the 2023 international guideline. A fertility clinic report reading “AFC: 34” is not the same kind of number as a diagnostic scan reporting “follicle number per ovary: 22” — one is a bilateral total built to plan a stimulation cycle, the other is a single-ovary count built to help diagnose PCOS, and mixing the two up is an easy, common mistake.

That difference exists because AFC was developed to answer a different question than the Rotterdam criteria: how ovaries are likely to respond to stimulation medication, not whether PCOS is present. A 2010 consensus effort focused specifically on standardizing the measurement found considerable variability in the clinical definitions and technical methods used to count and measure antral follicles across both research and routine practice — a variability problem that predates, and is separate from, the diagnostic-threshold question.

Why does the antral follicle count threshold depend on the ultrasound machine?

The diagnostic per-ovary threshold has already moved once, and specifically because equipment changed. A 2013 study of 168 women — 98 with PCOS by NIH criteria and 70 controls, all scanned transvaginally — found that newer, higher-resolution transducers required raising the research threshold to 26 follicles to keep correctly telling PCOS ovaries apart from typical ones, because better equipment could see follicles older machines missed entirely. The 2023 guideline eventually settled on 20 as its current figure, but it ties that number explicitly to a transducer running at 8 MHz or higher, rather than stating a count meant to apply regardless of equipment — and it says outright that ultrasound thresholds for PCOS should be revised again as technology keeps improving. That same resolution requirement is why transvaginal versus abdominal ultrasound matters for whether a count is reliable at all, separately from which threshold number applies to it.

That same machine-dependency runs through AFC as a fertility-treatment measurement, not only through the diagnostic count. The standardization review behind AFC identified several further sources of disagreement beyond raw resolution: whether counting is done from a single 2D sweep or a reconstructed 3D volume, how systematically the person doing the counting was trained, and which day of the cycle the scan happens on — a single dominant follicle above 10 mm can crowd out, or be mistaken for, the smaller follicles the count is actually measuring if the scan isn’t timed to the early follicular phase.

What does antral follicle count actually predict?

Antral follicle count predicts how a stimulated cycle is likely to respond with roughly 80% accuracy, but it does not meaningfully predict whether that cycle ends in a pregnancy. A pooled analysis of 28 study databases covering 5,705 women having IVF — a general fertility-clinic population, not restricted to PCOS, though PCOS is common among the high responders these tests are built to flag — found AFC alone reached an area under the curve of 0.76 for predicting a poor response to stimulation, close to AMH’s 0.78, with no real improvement when the two were combined (0.80). For predicting an ongoing pregnancy from that same treatment cycle, age alone reached only 0.57, and neither AFC nor AMH added anything to that number at all.

A related pooled analysis, focused on the opposite pattern — excessive response, the pattern more relevant to PCOS — found AFC performed with 82% sensitivity and 80% specificity for identifying who would over-respond to stimulation, comparable to AMH’s 82% sensitivity and 76% specificity, across a pooled set of studies using standard hyperstimulation in general IVF populations.

Read together, both analyses land on the same honest conclusion: AFC forecasts a cycle’s mechanics — roughly how many follicles are likely to grow, and how much monitoring a stimulation plan may need — not whether that cycle results in a pregnancy.

What does a high antral follicle count mean if you have PCOS?

PCOS ovaries carry roughly two to three times the typical number of small, early-stage follicles, a difference documented directly by comparing follicle counts between 59 people with PCOS and 45 controls — the same excess that makes AMH run high in PCOS, since AMH is produced by exactly these follicles. In practice, this means a fertility clinic scanning someone with PCOS is likely to see an AFC well above the range typical of the general IVF population the tables above were built from, which is a large part of why PCOS is a recognized risk factor for the excessive-response pattern those studies measure. A bigger starting follicle pool means a bigger stimulated response is more likely — not a diagnosis, and not a guarantee of an easier path to pregnancy.

Guidance for fertility clinics reflects this directly. The 2020 ESHRE guideline on ovarian stimulation recommends a risk assessment for high response before stimulation begins, using AFC or AMH — while also acknowledging that definitions of “high response” still vary enough across the published research that its own guideline group called them ill-defined even with an existing consensus definition in place. That inconsistency is worth knowing before treating any single AFC number as a hard line: what a higher count changes is the probability of a large response and the amount of monitoring a cycle may need, not a fixed category a person definitely falls into.

Is antral follicle count used to diagnose PCOS on its own?

No single antral follicle count diagnoses PCOS by itself. The diagnostic criterion is the per-ovary count covered above — polycystic ovarian morphology, one of three Rotterdam features — and meeting it never settles a diagnosis alone. Where AFC-style thinking does connect to diagnosis is through AMH: because AMH and follicle count track the same biology, the 2023 guideline allows an elevated AMH to substitute for the ultrasound leg of the criteria in adults, though never alongside ultrasound and never in place of the other two Rotterdam features. See AMH versus antral follicle count for PCOS diagnosis for how clinicians actually choose between the two.

Age is not a minor footnote here either. Ultrasound-based ovarian criteria, antral follicle count included, are not recommended for diagnosing PCOS within eight years of a first period, because multi-follicular ovaries are a normal, temporary feature of adolescent development rather than a diagnostic sign. The 2023 guideline excludes the ovarian criterion entirely at that life stage, whatever a scan shows.

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

Who this doesn’t help

An antral follicle count will not tell you whether a specific treatment cycle will end in a pregnancy — the 5,705-woman analysis above found that ovarian reserve tests, AFC included, added nothing to age’s already-limited ability to predict that outcome. It will not diagnose PCOS on its own, and it will not tell you your current egg quality, since follicle quantity and egg quality are measured separately and are not the same question. It is also not validated for diagnosis under 20, for the adolescent reasons covered above, whatever a teenager’s count shows.

A single AFC also doesn’t capture change over time by itself — if a follicle pool is rising or falling, whether from treatment or simply from age, that requires comparing counts taken the same way on different occasions, not reading one number alone. How ovarian reserve markers shift during PCOS treatment is answered on its own, separately from what a single AFC snapshot can tell you.

Your next step

If antral follicle count is part of your own fertility work-up, ask for the actual number, which ovary or ovaries it came from, and what cycle day and equipment the scan used. Those three details are what let a clinician read the result in context, rather than as a figure floating free of the conditions that produced it — and they’re the same details worth bringing to any second opinion or new clinic, since a different machine can return a different count from the same ovaries. Where that number fits inside the rest of a fertility evaluation is a conversation for the person who ordered it, not a threshold to check against alone — one piece of the wider fertility picture, rather than a verdict by itself.