Search this and you will get a list. Fifteen or twenty items, run together in one column — irregular periods, acne, weight gain, fatigue, hair growth, hair loss, mood, bloating, cravings — with the implication that the more of them you recognise, the more likely it is that you have found your answer.
That is not how the diagnosis works, and the list format actively obscures how it does. Three findings are criteria. Everything else is associated. Sorting your own symptoms into those two groups is more useful than adding to the list, because only one group is what a clinician is assessing against.
The three that are criteria
PCOS is diagnosed against the Rotterdam criteria — two of three, after conditions that mimic it have been excluded.
| Criterion | How it shows up |
|---|---|
| Ovulatory dysfunction | Irregular or absent cycles — and “irregular” has a definition with numbers that changes with years since your first period |
| Hyperandrogenism | Clinical: hirsutism most reliably, with acne and scalp hair loss less specific. Or biochemical: raised androgens on bloods |
| Polycystic ovarian morphology | 20 or more follicles in one ovary on ultrasound, or a raised AMH in adults — not in adolescents |
Three things about that table do most of the work.
Two is enough, and three is not required. Someone with regular cycles can still meet the criteria. So can someone with a normal ultrasound.
The second criterion has two routes. Hyperandrogenism can be met clinically or biochemically, which is why normal bloods do not close the question for someone with visible signs.
The third has an age rule. Neither ultrasound nor AMH is used for diagnosis within eight years of a first period, and adolescent diagnosis works differently — both of the other findings required, rather than two of three.
The ones that are real but are not criteria
These are common, they are documented, and several of them are the reason people go to a doctor in the first place. None of them counts towards the two-of-three.
Metabolic features. Insulin resistance is central to the condition and is not a diagnostic criterion — one of the stranger facts about the whole subject, and the reason the guideline asks for glycaemic screening in everyone with the diagnosis regardless of weight.
Fatigue. Very commonly reported, and the symptom most likely to be attributed to the diagnosis rather than investigated. The guideline asks a specific screening question about sleep that most people are never asked.
Weight changes. Associated, and not a criterion. PCOS occurs across the whole weight range, and the associated insulin resistance is found independently of body weight — which is why a weight-centred account is not the whole account.
Mood and quality of life. The guideline treats emotional wellbeing as part of the picture rather than a side issue. It is almost never in the ten-minute version.
Bloating, cravings, cycle-related pain. Frequently reported, not criteria, and pain in particular is worth raising on its own terms rather than being absorbed into an existing diagnosis — it is the dominant feature of a different condition that is often present alongside.
Why the distinction is worth making
It changes what you can ask for.
If your symptoms are mostly in the second group, the useful question is not “do I have PCOS” but whether the criteria have actually been assessed — which cycle history was recorded, whether androgens were measured and by which method, whether the third criterion was addressed at all. Those are answerable.
It also protects against the opposite error. Recognising twelve items on a list and concluding you have the condition is how people arrive at an appointment with a conclusion rather than a history, which reliably goes worse. Several conditions produce this picture, and the guideline asks for them to be excluded first — not as a formality, but because the alternative is treating the wrong thing.
A list is not a diagnosis
This page cannot tell you whether you have PCOS and neither can any other page. The criteria are specific, they require exclusions, and they are assessed by someone who can order the tests and read the results against the right ranges.
What a list of signs is genuinely good for is deciding that it is worth being assessed. That is a lower bar than it sounds and it is the correct use of it.
What to do with the ones you recognise
Record them, in the form the criteria are made of.
The first criterion is dates — the day bleeding started, nothing more complicated. The second is a trend: skin and hair change slowly, and the comparison being asked for at an appointment is with a version of yourself you stopped noticing. The third is a test, and nothing you do changes it.
Then take the record rather than the recollection. What to log, and the much shorter list that survives a bad month, is its own question — and the answer is not “everything”, because a record that collapses under its own weight tells nobody anything.