Of the three findings that can add up to PCOS, this is the one most people picture when they imagine being tested for it. It is also the one whose definition has moved, and whose old definition is still being quoted on most of the pages that explain it.
The number is 20. A great deal of what you will read says 12.
What is actually counted
The scan is looking for polycystic ovarian morphology — the third of the Rotterdam criteria. In the 2023 international guideline, the threshold in adults is:
| Measure | Threshold | When it is used |
|---|---|---|
| Follicle number per ovary | 20 or more in at least one ovary | The primary measure, with current equipment |
| Ovarian volume | 10 ml or more in at least one ovary | Older technology, or image quality insufficient for a count |
| Follicle number per section | 10 or more in at least one ovary | Same situation as above |
One ovary meeting a threshold is enough; both are not required. And the alternatives are not lesser versions of the same thing — they exist because a follicle count depends on being able to resolve follicles, and when the equipment or the image cannot, counting them produces a confident number that is not reliable.
Why 12 is still everywhere
Twelve was the Rotterdam threshold set in 2003, on the imaging equipment of 2003. Transducers improved. A machine that resolves more follicles finds more follicles in the same ovary, and a threshold calibrated to older equipment, applied to newer equipment, starts classifying ordinary ovaries as polycystic.
Raising the threshold was not a change of opinion about what PCOS is. It was a correction for the instrument having got better — the same measurement, recalibrated.
The practical consequence is worth stating plainly, because it is the reason this page exists: a count of 15 follicles is below the current threshold and above the one most of the internet is quoting. Pages written before 2018 and never revised, and newer pages copied from them, will tell someone in that position that their ovaries are polycystic. The current guideline does not.
What the report should say
The guideline sets minimum reporting standards, and they are more interesting than they sound. A report should include the last menstrual period, the transducer bandwidth frequency, and the approach or route used.
Those are not clerical details. The transducer frequency is what determines which threshold applies — it is the difference between a count that can be read against 20 and an image that should be reported on volume instead. The route matters for the same reason. And the cycle day affects what is there to see.
Which gives a reasonable question for anyone holding a report with a number on it and no context: which approach was used, and at what frequency? The answer determines how much weight the count can carry, in the same way that the assay method determines what a hormone result means.
Transvaginal, and the word “acceptable”
The guideline describes transvaginal imaging as the most accurate approach where it is acceptable to the patient.
That second clause is doing real work and is usually dropped in summary. It frames the scan as a choice, with a transabdominal alternative that is reported differently — generally on ovarian volume rather than a follicle count — rather than as a requirement to be endured.
There is also now a route that avoids the question entirely for some people. In adults, a raised AMH can stand in for the scan for this criterion. That option did not exist before 2023 and it is not widely known.
Who should not have this scan for diagnosis
The guideline gives no ultrasound criteria for adolescents and does not recommend the scan for diagnosis in that group.
Multifollicular ovaries are common in healthy adolescents, so the scan does not separate PCOS from ordinary development — it mostly identifies being young. Neither ultrasound nor AMH is recommended within eight years of a first period, which is the whole shape of adolescent diagnosis: both other findings required, and the imaging criterion removed rather than relaxed.
What most articles get wrong
The number, as above. Twelve is a 2003 figure and it is the single most repeated error on this subject.
They call the follicles cysts. They are not cysts. They are ordinary follicles, each containing an immature egg, and every ovary has them — the finding is about how many are visible, not about anything having gone wrong structurally. The condition’s name is a historical misnomer that the guideline’s own authors have argued about publicly.
They treat the scan as the test. It addresses one criterion of three, a diagnosis needs two, and conditions that mimic PCOS have to be excluded first. A scan alone cannot diagnose PCOS and a normal scan cannot exclude it.
What this changes at an appointment
Three things are worth carrying in.
If you have a report with a follicle count, read it against 20. Then check which threshold applies, by asking about the approach and the transducer, because a volume-reported scan is answering a different question.
If a scan has been proposed and you would rather not have one, the guideline’s own wording — most accurate where acceptable — plus the AMH route in adults makes that a reasonable conversation rather than a refusal.
If the scan is normal and your symptoms are not, that closes one criterion and nothing else. Two of three is the rule, and the other two are a cycle history and hyperandrogenism, neither of which a sonographer can see.