The diagnostic rule for PCOS is short enough to fit in one line: two of three findings, with other explanations excluded. Almost every confusion about this condition comes from that structure, so it is worth understanding properly rather than looking up which two you have.
The three are:
- Ovulatory dysfunction — irregular, infrequent or absent ovulation
- Hyperandrogenism — raised androgens, measured in blood or visible as symptoms
- Polycystic ovarian morphology — the ovarian appearance the condition is named after
Any two make the diagnosis. Which means there are three different ways to have PCOS, and someone diagnosed on 1 and 2 may have nothing in common, symptomatically, with someone diagnosed on 2 and 3.
Criterion one: ovulatory dysfunction
This is usually the one people notice first, because it shows up as periods that are late, absent, or wildly unpredictable.
Two things are worth separating here. A period is not the same as ovulation — you can bleed without having ovulated, which is why “I still get periods” does not settle the question. And an occasional long cycle is not the criterion; what is being looked for is a persistent pattern.
The practical consequence: this criterion is established from dates, over months. It is the one you can contribute to most directly, because nobody else is in a position to know when your last four periods started. Keeping that record when the cycles are irregular takes some care, because the ambiguous months are the informative ones.
Criterion two: hyperandrogenism
Androgens are a group of hormones, testosterone among them, present in everyone and higher in PCOS in a large proportion of cases.
This criterion can be met two different ways, and this is where most misreadings happen:
- Biochemical — a raised measurement on a blood test. Usually total testosterone alongside SHBG, from which a free androgen index is calculated.
- Clinical — visible signs, without any blood test needing to agree. Hirsutism is the most specific of these; acne and scalp hair thinning also count, though both are common enough on their own to be weaker evidence.
Either satisfies the criterion. So a normal testosterone result does not rule out hyperandrogenism, and this is the single most common misreading of a PCOS blood panel — what each of those tests is actually measuring is worth knowing before the results conversation.
Criterion three: polycystic ovarian morphology
The most misleading part of the name. The “cysts” are not cysts in the sense anyone means by that word — they are ordinary small follicles, present in every ovary, simply more numerous. Nothing needs draining and nothing is growing.
Two facts follow, and both matter:
Polycystic-looking ovaries are common in people without PCOS. Which is exactly why this cannot be the diagnosis on its own, and why a scan report saying “polycystic ovaries” is a finding rather than an answer.
You can have PCOS with entirely normal-looking ovaries. Criteria 1 and 2 together are a complete diagnosis. This surprises people who were told they could not have it because a scan came back clear.
What changed in 2023
The 2023 update of the international guideline made two changes here that are worth knowing about.
An AMH blood test can now replace the ultrasound in adults. Anti-Müllerian hormone rises roughly with the number of small follicles, so it measures the same underlying thing by a different route — and a blood test is available in primary care in a way a gynaecological ultrasound often is not.
Neither is needed when the first two criteria are already met. If cycles are irregular and androgens are raised, the diagnosis is complete on two criteria and a scan adds nothing. If you were sent for one anyway, it may have been for another reason — but it was not needed to answer this question.
In adolescents, neither ultrasound nor AMH is recommended for this at all. In the years after menarche, polycystic-looking ovaries and irregular cycles are both common enough that neither distinguishes much. For adolescents the guideline requires both hyperandrogenism and ovulatory dysfunction — a deliberately higher bar, to avoid diagnosing a syndrome from what is a normal phase for a lot of people.
The part of the rule nobody quotes
“Two of three” is only half the criteria. The other half is exclusion of conditions that produce the same picture, and it is the reason a PCOS blood panel contains tests that appear to have nothing to do with ovaries.
Thyroid disease, raised prolactin and non-classic congenital adrenal hyperplasia can all cause irregular cycles, raised androgens, or both. They are managed differently. So TSH, prolactin and 17-hydroxyprogesterone are usually measured not to confirm PCOS but to rule out something else, and their being normal is part of why the answer was PCOS.
If you have wondered why the panel was so long, this is the reason.
Why two people with the same diagnosis look nothing alike
Because the criteria are a menu, not a checklist.
Someone with irregular cycles and hirsutism has PCOS. Someone with regular-ish cycles, raised androgens and polycystic ovaries has PCOS. Someone with absent periods and polycystic ovaries and no visible androgen signs has PCOS. Those are three genuinely different presentations under one name.
Which explains a lot of the frustration in PCOS communities — advice that transformed one person’s experience does nothing for the next, and neither of them is doing it wrong. It also explains why “do I really have it?” is such a persistent question: if your version is missing the feature you assumed was definitional, the diagnosis feels unearned.
It is not. Two of three is the rule, and there is no bonus criterion for having the third.
What this means for your own record
The diagnosis is made once. What happens afterwards is managed on change over time — cycles, symptoms, metabolic markers — and none of that is visible in a single appointment.
That is the practical case for keeping the record yourself. A diagnosis established from two criteria says what you have. A year of dated entries says what it is doing, which is the question every appointment after the first one is actually about.