How to Know If You Have PCOS: How Doctors Diagnose Polycystic Ovary Syndrome

Short Answer

There is no single "PCOS test" you can take at home or in one blood draw. Doctors diagnose PCOS by putting several pieces together and first making sure something else is not driving the same pattern. In adults, the diagnosis is usually made when at least two of three Rotterdam features are present: irregular or absent ovulation, signs of high androgens in your body or on bloodwork, and polycystic ovarian morphology on ultrasound. Before calling it PCOS, a clinician also needs to rule out common look-alikes such as thyroid disease, high prolactin, and non-classic congenital adrenal hyperplasia. (Endocrine Society guideline)

The 2023 International Evidence-Based Guideline keeps that same "two of three" structure and adds one important update for adults: anti-Müllerian hormone, or AMH, can sometimes be used instead of ultrasound to help define polycystic ovarian morphology. But AMH is not a stand-alone diagnosis. A single high AMH result, a home hormone kit, or an app pattern cannot tell you "yes, this is PCOS" without the clinical context — your cycle pattern, androgen symptoms or lab results, age, medications, and exclusion of other causes all matter. (2023 International Evidence-Based Guideline)

Welltory does not measure hormones and is not a diagnostic test. What it can do is help you bring a clearer timeline to your clinician: cycle regularity, sleep changes, resting heart rate, HRV patterns, stress load, and symptoms you've noticed around them. That record cannot diagnose PCOS, but it can make the medical conversation less vague.

PCOS diagnosis at a glance

PCOS is usually diagnosed with the Rotterdam framework: in adults, doctors look for at least two of three patterns — ovulatory dysfunction, hyperandrogenism, and polycystic ovarian morphology — and they make the diagnosis only after ruling out other conditions that can look similar. (2023 International Evidence-Based Guideline (Hum Reprod))

The key point: PCOS is not diagnosed from one symptom, one scan, or one hormone result. The pieces have to line up — at least two Rotterdam features — and your clinician also has to exclude common mimics such as thyroid disease, high prolactin, and nonclassic congenital adrenal hyperplasia before calling it PCOS. (Endocrine Society guideline)

Tests used in a PCOS work-up

There isn't one "PCOS test." The work-up is a pattern check: your clinician looks for ovulation problems, signs or blood-test evidence of higher androgens, and — in adults when needed — polycystic ovarian morphology on ultrasound or AMH. Just as important, they rule out look-alike conditions before putting the PCOS label on your chart. (2023 Guideline (Hum Reprod))

PCOS is a diagnosis of exclusion, not a single test

The first thing to understand is that PCOS is not confirmed by one blood draw, one scan, or one home hormone kit. Doctors diagnose it by matching your pattern to a defined framework — the Rotterdam criteria — and by checking that something else is not creating the same signs in your body. That second part matters. Irregular or absent periods, acne, new facial or body hair growth, hair thinning, or trouble conceiving can come from PCOS, but they can also come from thyroid disease, high prolactin, nonclassic congenital adrenal hyperplasia, or, in some situations, other hormone conditions. So a careful work-up usually starts with your cycle history and signs of androgen excess, then may add blood tests and pelvic ultrasound; it also checks the "look-alikes" before anyone writes down "PCOS" as the diagnosis. (2023 Guideline (Hum Reprod))

Under Rotterdam, an adult diagnosis needs at least two of three features: irregular or absent ovulation; hyperandrogenism, meaning visible signs such as excess hair growth or acne and/or elevated androgens on blood testing; and polycystic ovarian morphology on ultrasound. Because only two of the three are required, PCOS does not look the same in every person. One person may have irregular cycles and high androgens without needing ultrasound to make the diagnosis; another may have ovulation problems and polycystic-appearing ovaries without obvious androgen symptoms. That heterogeneity is real, not "in your head." Recent phenotyping research underscores it: one study found that "the HA phenotype was characterised by significantly more severe oligo-anovulation and metabolic disturbance" than the non-hyperandrogenic form, and concluded that "insulin resistance drives the hyperandrogenic phenotype independently of obesity." (Phenotyping study, Front Endocrinol)

The three Rotterdam criteria, one at a time

Before the three pieces, there's one rule that matters: in adults, PCOS is usually diagnosed when at least two of these three criteria are present and other conditions that can look like PCOS have been ruled out. That's why the appointment may feel broader than "checking for PCOS" — your clinician is also checking whether thyroid disease, high prolactin, congenital adrenal hyperplasia, or another cause better explains the pattern. (2023 Guideline (Hum Reprod))

Ovulation. Irregular, infrequent, or absent periods are often the entry point because they can mean ovulation is not happening regularly. Your clinician will usually map your cycle length and pattern over time — how many days between bleeds, whether you sometimes go months without a period, and whether the pattern has been persistent. This part is mostly history, not a single lab result, which is why a cycle record can be genuinely useful: it gives your doctor the pattern your ovaries have been showing across real life, not just on one day in the clinic. (2023 Guideline recommendations)

Androgens. High androgens can count in two ways. They can be clinical, meaning they show up on your body as hirsutism, persistent acne, or androgen-pattern scalp hair thinning. Or they can be biochemical, meaning a blood test shows elevated androgens — usually total testosterone and free testosterone or free androgen index first, with androstenedione or DHEA-S sometimes considered if testosterone results do not explain the picture. Either route can satisfy the hyperandrogenism criterion; you do not need both visible symptoms and abnormal labs for this piece to count. (Endocrine Society guideline)

Ovarian morphology. Ultrasound looks at ovarian structure — mainly follicle number and ovarian volume — but this result is not interpreted in isolation. In adults, polycystic ovarian morphology can be one of the three Rotterdam criteria. In adolescents and people who are still close to their first period, it is different: guidelines advise not using ultrasound morphology to diagnose PCOS within 8 years after menarche, because multi-follicular ovaries are common in normal puberty and the finding is not specific enough. In that window, doctors lean more heavily on persistent ovulatory dysfunction plus clinical or biochemical hyperandrogenism, with follow-up over time if the picture is incomplete. (Adolescent PCOS guidance, BMC Medicine)

Where AMH fits — and why it is not a "PCOS test"

AMH — anti-Müllerian hormone — is made by cells around early-growing ovarian follicles, so it tends to rise when the ovaries contain many small follicles, as often happens in PCOS. That is why AMH now has a legitimate place in the diagnostic work-up. In adults, the 2023 International Evidence-Based Guideline allows serum AMH to be used instead of ultrasound to define polycystic ovarian morphology, which can make the process simpler when ultrasound is difficult, unavailable, or not needed. But the same guideline is clear about the limit: AMH should be used inside the diagnostic algorithm, not as a stand-alone answer, and it is not recommended for diagnosing adolescents. (AMH in PCOS review)

That distinction matters because PCOS is a pattern, not one lab number. If you already have irregular ovulation plus clinical or biochemical signs of excess androgens, your clinician may not need AMH at all to make the diagnosis. If the picture is less clear, AMH can help stand in for the ovarian-morphology part of the criteria — but your age, BMI, recent hormonal contraception, ovarian surgery history, menstrual-cycle timing, and the specific lab assay can all shift the result. A high AMH can support the story your body is telling; it cannot tell the whole story by itself. (2023 Guideline (Hum Reprod))

The modeling data say the same thing in a different language. In one study, "anti-müllerian hormone alone predicted PCOS diagnosis with relatively good diagnostic accuracy (AUROC 0.884)," with an F1 score of 0.807 — good, but not definitive. When AMH was paired with ovarian volume, performance improved: "integration of AMH and ovarian volume improved model performance (AUROC 0.906)," with an F1 score of 0.811. In other words, AMH became more useful when it was interpreted alongside another sign from the ovaries, not when it was treated like a yes/no PCOS switch. (Predictive modeling study, O&G Open)

The control group in that study was deliberately strict — "controls had no diagnostic features of PCOS" — because real life is messier. AMH values can overlap between people with and without PCOS, and different PCOS phenotypes do not all look the same on labs. The authors' practical takeaway was not "AMH diagnoses PCOS." It was that "a minimum combination of ovarian volume with AMH and examination/history-derived end points can accurately predict PCOS diagnosis." So if you see an AMH result on a portal or a home hormone panel, bring it to your clinician — along with your cycle pattern, acne or hair-growth changes, weight and metabolic history, and any prior ultrasound. AMH is a clue. Diagnosis still comes from the whole pattern. (Predictive modeling study, O&G Open)

Ruling out the look-alikes (why exclusion matters)

PCOS is not diagnosed by spotting one symptom and naming it. It is diagnosed after your clinician checks whether the same pattern could be coming from somewhere else in the hormone system. Thyroid disease can disrupt ovulation and make cycles irregular. High prolactin can stop or delay periods. Nonclassic congenital adrenal hyperplasia can raise adrenal androgens and look like "PCOS-type" acne, hair growth, or cycle changes. That is why guideline-based work-ups usually include thyroid-stimulating hormone (TSH), prolactin, and 17-hydroxyprogesterone (17-OHP) before the PCOS label is applied; the Endocrine Society guideline specifically names thyroid disease, hyperprolactinemia, and nonclassic congenital adrenal hyperplasia screened primarily with serum 17-OHP as conditions to exclude, and notes that women with PCOS-like signs should be screened with TSH, prolactin, and 17-OHP. (Endocrine Society guideline)

This exclusion step also appears in newer diagnostic frameworks. In a nationwide prevalence survey using the 2023 framework, the researchers describe screening women "utilizing the 2023 International Evidence-Based Guidelines, incorporating anti-Mullerian hormone," and note that "thyroid dysfunction and hyperprolactinemia were excluded" before a diagnosis was made. The 2023 international guideline recommendations also keep the same principle: other causes should be excluded, and the diagnostic algorithm lists TSH, prolactin, 17-OH progesterone, FSH, and clinically indicated evaluation for conditions such as Cushing's syndrome or adrenal tumors. (Nationwide prevalence survey)

The same survey is a reminder that a partial picture is not a diagnosis. One clue — irregular cycles, acne, extra hair growth, a high androgen result, or a high AMH result — can matter, but it does not automatically explain the whole body. Many women had one feature but not enough to confirm PCOS, and the authors concluded that "women with isolated diagnostic criteria require further evaluation and long-term follow-up." In practice, that means your doctor is looking for a pattern over time: what your cycle does, whether androgen signs persist, what blood tests show, and whether another condition explains it better. (Nationwide prevalence survey)

Can a home test or wearable diagnose PCOS?

No. Direct-to-consumer hormone tests — including at-home AMH or androgen kits — can give you a lab value, but PCOS is not a single-number diagnosis. Doctors diagnose it by fitting the whole picture together: cycle pattern, signs or lab evidence of excess androgens, ovarian morphology assessed by ultrasound or AMH within the diagnostic algorithm, and the exclusion of look-alike conditions. (2023 Guideline (Hum Reprod))

AMH is the clearest example of why context matters. The 2023 International Evidence-based Guideline says serum AMH can be used to help define polycystic ovarian morphology in adults, but it should be used only within the diagnostic algorithm and should not be used as a single test for PCOS. So a high AMH does not automatically mean you have PCOS, and a normal AMH does not safely rule it out. FDA guidance makes the same broader point about direct-to-consumer and home-use tests: results should not replace a traditional health care evaluation, and many tests are best interpreted alongside your medical history, physical exam, and other testing. (2023 Guideline (Hum Reprod); FDA: Direct-to-Consumer Tests)

Welltory is not a diagnostic test and does not measure hormones. What it can do is help you see patterns over time — cycle-length changes, sleep disruption, resting heart rate trends, and HRV shifts — so you can bring a clearer timeline to your appointment. That record can make the conversation with your clinician sharper and less dependent on memory. It still does not diagnose PCOS, rule it out, or replace medical judgment.

Who diagnoses PCOS — and why it often takes time

PCOS is most often worked up by a gynecologist, reproductive endocrinologist, or endocrinologist, but your primary-care clinician can start the process: taking your cycle history, checking acne or excess hair growth, ordering initial blood tests, and deciding whether you need referral. There isn't one "PCOS test." Doctors piece together a pattern — ovulation problems, signs of higher androgens, and ovarian findings — while also checking for conditions that can look similar, such as thyroid disease, high prolactin, or adrenal causes. That's why the work-up may include a physical exam, hormone blood tests, metabolic labs, and sometimes pelvic ultrasound rather than a single yes/no result. (Mayo Clinic: PCOS diagnosis)

This is also why diagnosis can take more than one visit. Your body may not show the whole pattern at once: periods can be irregular for months and then temporarily settle; acne or hair growth may appear gradually; lab results can be borderline; and ultrasound findings do not tell the whole story on their own. In a large international survey of women with PCOS, 33.6% reported that it took more than 2 years to get a diagnosis, and 47.1% reported seeing 3 or more health professionals before the diagnosis was established. (Gibson-Helm et al., J Clin Endocrinol Metab)

So if it took a while to get answers, that does not mean you missed something obvious or failed to explain yourself well enough. It reflects how PCOS is diagnosed: by matching a changing symptom pattern to criteria, then ruling out other hormone problems that can send the body the same signals. A good clinician should make the next step clear — what has been ruled out, what still needs checking, and when to follow up if you only meet part of the picture right now.

When to see a doctor

See a clinician if your periods are consistently irregular or absent, if you're dealing with excess facial or body hair, persistent acne, scalp hair thinning, trouble getting pregnant, or signs of insulin resistance such as dark, velvety skin patches on the neck, armpits, groin, or under the breasts. These symptoms can feel like separate problems — skin, hair, cycles, weight, fertility — but in PCOS they can come from the same hormone-and-ovulation pattern, so it's worth getting them checked rather than waiting for the picture to become "obvious." (Mayo Clinic: PCOS symptoms)

Also book an appointment if the symptoms are affecting your daily life, work, relationships, or mental health. A clinician can look for PCOS and rule out other causes that can mimic it, then decide whether you need hormone blood tests, metabolic screening, a pelvic exam, ultrasound, or referral to a gynecologist, endocrinologist, or fertility specialist. (NHS: PCOS diagnosis)

Seek prompt care for very heavy, prolonged, or otherwise unusual vaginal bleeding. Heavy bleeding can have many causes — some hormonal, some not — and it shouldn't be assumed to be "just PCOS." (Mayo Clinic: vaginal bleeding)

Before the visit, bring whatever pattern data you have: cycle dates, how long bleeding lasts, how heavy it is, when acne or hair changes started, weight or sleep changes, medications and supplements, and whether you're trying to conceive. A simple timeline helps your clinician connect the dots faster and makes the first appointment more productive. (Mayo Clinic: PCOS diagnosis)

How we made it

Made with AI tools, then edited, fact-checked, and medically reviewed by the Welltory team. Any cohort figures are reported only as anonymized, aggregated data; no individual user is identifiable.

Discounts for blog readers: up to 36% off

See what affects your energy, stress, sleep, and daily state with Welltory

Explore your data