How PCOS Is Tested and Diagnosed: Blood Tests, Ultrasound and the Rotterdam Criteria

The appointment that finally explains things often starts with a calendar. A woman flips back through her phone, counting: a period in January, nothing until late March, then two in May. Her doctor nods, asks about hair on her chin, glances at the acne along her jaw, and says the sentence she has half-expected for years. That conversation, more than any machine, is where a PCOS diagnosis begins.

Polycystic ovary syndrome is one of the most common hormonal conditions in people of reproductive age, affecting roughly 1 in 10 women in the UK according to the NHS, and it is also one of the most confusing to have confirmed. There is no single blood marker. The scan that gives the condition its name is optional. And the label itself is misleading, because the “cysts” are not cysts at all.

What follows is a clear-eyed walk through how clinicians actually reach the diagnosis: which questions matter, which blood tests are ordered and why, what the ultrasound can and cannot tell you, and where the process commonly stalls.

Why is PCOS diagnosed by pattern rather than a single test?

Most people arrive expecting a yes-or-no result, the way a pregnancy test or a strep swab works. PCOS does not behave that way. It is a syndrome, a recognizable cluster of features that tend to travel together, rather than a disease with one identifiable cause and one measurable signature.

The underlying biology helps explain the fuzziness. In PCOS, the ovaries produce more androgens (the so-called male hormones that everyone makes in small amounts) than usual, and the brain’s signals to the ovary become mistimed, so a dominant follicle often fails to mature and release an egg. Many people also have insulin resistance, which nudges the ovaries toward yet more androgen production. Each of those threads can be measured, but no single measurement is abnormal in every person with the condition.

Consider two patients. One has clockwork-regular periods but persistent chin hair and a testosterone level above the reference range. Another has three periods a year, normal testosterone, and ovaries carrying dozens of small follicles on ultrasound. Both can meet the accepted definition of PCOS; neither would be caught by a rule that demanded one particular finding.

That is why the diagnosis, in the framing used by the NHS, Mayo Clinic and the NIH, rests on combining history, physical examination, targeted blood work and, sometimes, imaging, while actively excluding other conditions that can mimic the picture. The thoroughness can feel slow. It is also what prevents someone with an undiagnosed thyroid problem from being given the wrong label for a decade.

What are the Rotterdam criteria for PCOS?

The framework most clinicians use was agreed at a consensus meeting in Rotterdam in 2003 and has been refined since. It asks whether a person has at least two of three features, once look-alike conditions have been excluded.

• Ovulatory dysfunction: infrequent, irregular or absent periods, reflecting cycles in which an egg is not reliably released.
• Hyperandrogenism: either visible signs such as excess facial or body hair, persistent acne or scalp hair thinning, or a blood test showing elevated androgens.
• Polycystic ovarian morphology: ovaries on ultrasound that are enlarged or contain an unusually high number of small follicles.

Two of three is the key phrase. Someone with irregular cycles and high testosterone meets the definition without ever having a scan, which is why the NHS and Mayo Clinic both note that ultrasound is not always needed. Equally, a person with normal hormone tests can still qualify if their cycles are irregular and their ovaries show the typical appearance.

Because the criteria allow several combinations, PCOS is sometimes described in phenotypes: all three features, or two of the three in various pairings. The distinction matters for research and can shape which long-term health checks a clinician prioritizes, but the diagnosis itself is the same.

The phrase “once other causes are excluded” is not fine print. It is the reason the first round of blood tests often has more to do with the thyroid and pituitary gland than with the ovaries, a point we will come back to.

What counts as an irregular period when testing for PCOS?

“Irregular” is one of those words that means something different to every person who uses it. A cycle that arrives every 26 days one month and 31 the next is variable but normal. Clinicians are looking for something more specific.

Mayo Clinic describes the pattern as having fewer than nine periods a year, or cycles that run more than 35 days apart. Bleeding that lasts many days longer than usual, or unpredictable heavy bleeding after long gaps, also raises the question. The common thread is anovulation: without a released egg, the hormonal cascade that triggers a timely period does not happen, so the lining builds until it sheds on its own schedule.

Two subtleties trip people up. First, someone taking hormonal contraception will usually have regular withdrawal bleeds regardless of what their ovaries are doing, so those bleeds tell the clinician very little. Your doctor will want to know what your cycles were like before you started, even if that was years ago. Second, cycles that look regular can occasionally be anovulatory; if the picture is otherwise suggestive, a blood test timed to the second half of the cycle can check whether ovulation actually occurred.

Bring data if you have it. A year of dates from a tracking app, or even a rough tally of how many periods you had last year, is genuinely useful. The number of cycles per year is one of the few places in a PCOS workup where the patient’s own record is the primary evidence.

What does a doctor look for at the first PCOS appointment?

Before anyone draws blood, a good consultation does most of the diagnostic work. Expect questions that range wider than your periods: when they started, how they have behaved since, whether you have tried to conceive, what medicines and supplements you take, how your weight has changed, whether close relatives have diabetes or similar symptoms.

The physical examination is deliberately attentive to skin and hair. Clinicians look for terminal hair (coarse, pigmented strands) in a male-pattern distribution: upper lip, chin, chest, lower abdomen, inner thighs. Standardized scoring systems exist, but the impression of someone who has done hundreds of these exams is often just as informative. Persistent acne beyond the teenage years, thinning at the crown or temples, and velvety darkened skin at the neck or armpits (a marker of insulin resistance called acanthosis nigricans) all go into the picture.

Blood pressure, height and weight are recorded, not to judge but because cardiovascular and metabolic risk shape which follow-up tests are sensible. Mayo Clinic lists a pelvic exam among the possible steps, though many clinicians reserve it for situations where another cause of symptoms needs checking.

Say the awkward things out loud. Hair removal habits matter, because someone who waxes or lasers weekly may show no visible hair at all. Skin treatments can mask acne. Doctors are not scoring your grooming; they are trying to reconstruct what your body does on its own.

Which blood tests are used to test for PCOS?

There is no “PCOS test” on a laboratory form. Instead, clinicians order a panel that answers two questions at once: are androgens raised, and is something else causing the symptoms? The table below summarizes what is commonly requested and why, drawing on the test lists published by the NHS, Mayo Clinic and Cleveland Clinic.

Test	What it measures	Why it is ordered
Total testosterone	Overall androgen level	Biochemical evidence of hyperandrogenism
SHBG and free androgen index	Protein that binds testosterone; calculated unbound fraction	Low SHBG (common with insulin resistance) can make normal total testosterone misleading
LH and FSH	Pituitary signals to the ovary	Supportive pattern; also screens for other causes of absent periods
Prolactin	Pituitary hormone	High levels can cause irregular periods and must be excluded
Thyroid function (TSH)	Thyroid activity	Under- or overactive thyroid mimics PCOS symptoms
17-hydroxyprogesterone	Adrenal steroid precursor	Screens for a form of congenital adrenal hyperplasia
Glucose, HbA1c or glucose tolerance test	Blood sugar handling	Insulin resistance and diabetes risk
Lipid profile	Cholesterol and triglycerides	Cardiovascular risk assessment

A few points deserve emphasis. Testosterone assays vary between laboratories, and reference ranges for women sit near the lower limit of what many assays measure well, so a single borderline value rarely settles anything. Anti-Müllerian hormone (AMH) is often high in PCOS and is being studied as a possible substitute for ultrasound, but international guidance has not yet endorsed it as a standalone diagnostic criterion, so treat any AMH result as a supporting clue rather than a verdict.

When should PCOS blood tests be taken, and what can skew them?

Timing is where a technically correct blood test can still give a misleading answer. Reproductive hormones swing across the menstrual cycle and across the day, so laboratories interpret them against the moment they were drawn.

If you are having periods, even infrequent ones, clinicians generally prefer to test in the early follicular phase, the first few days after bleeding begins, when LH, FSH and estrogen are at a predictable baseline. If your cycles are so irregular that you cannot plan, a random sample is acceptable; the results are simply read with that in mind. Testosterone follows a daily rhythm and is typically highest in the morning, which is why early appointments are often suggested.

The biggest confounder is hormonal contraception. Combined methods raise SHBG and suppress ovarian androgen production, so testosterone can read as normal in someone whose untreated level would be high, and the regular bleeds erase the cycle history. Your clinician may suggest a pause before testing, or may decide the clinical signs are strong enough to diagnose without it. That is a shared decision, weighing contraceptive needs against diagnostic clarity, and it belongs with the prescribing clinician.

Other things that muddy the water include recent significant weight change, very intense exercise, high stress or illness, all of which can suppress ovulation for reasons unrelated to PCOS. Some supplements marketed for hormones can also shift results. Tell the person ordering the tests about everything you take, prescribed or not, so the numbers can be interpreted rather than simply recorded.

What does a PCOS ultrasound show, and do you need one?

The ultrasound is the part of the workup people picture most vividly and understand least. Start with the vocabulary: the “cysts” in polycystic ovaries are follicles, the small fluid-filled sacs that each contain an immature egg. The NHS describes them as harmless and typically no more than about 8 mm across. In PCOS, many of them stall at this size rather than one maturing and releasing, so the ovary looks like a string of pearls around its rim.

A sonographer counts follicles in each ovary and measures ovarian volume. The Rotterdam definition sets numeric thresholds for both; the follicle-count cutoff has been revised upward as scanner resolution improved, so the radiology report will apply the threshold appropriate to the equipment used. If the count or volume in either ovary exceeds it, the ovary is described as polycystic in appearance.

Transvaginal scanning gives the clearest images because the probe sits close to the ovaries. For people who have not had penetrative sex, or who prefer it, a transabdominal scan through a full bladder is used instead, accepting slightly lower precision.

Do you need the scan at all? Often not. Both the NHS and Mayo Clinic state that if irregular cycles and evidence of high androgens are already present, the criteria are met and imaging adds nothing to the diagnosis. Clinicians still request it in some cases to check for other causes of symptoms, to assess the uterine lining after long gaps between periods, or when only one criterion is present and the scan would decide the question.

Why does the doctor test your thyroid and other hormones first?

Here is the part of the Rotterdam criteria that gets skipped in casual summaries: PCOS is, formally, a diagnosis of exclusion. Several other conditions produce irregular periods, excess hair or acne, and a few of them need entirely different management. Missing them is the real diagnostic risk, far more than missing PCOS itself.

An underactive thyroid slows metabolism and can lengthen or halt cycles; an overactive one can do the reverse. A single TSH measurement usually settles the question. Raised prolactin, the hormone that drives milk production, suppresses ovulation and can cause unexpected breast discharge; it is sometimes traced to a benign pituitary growth or to certain medicines. Non-classic congenital adrenal hyperplasia, an inherited enzyme quirk, produces excess androgens from the adrenal glands rather than the ovaries and is screened for with a 17-hydroxyprogesterone level.

Rarer possibilities sit further down the list. Cushing’s syndrome, driven by excess cortisol, can bring weight gain, irregular periods and hair growth together. Androgen-secreting tumors of the ovary or adrenal gland are uncommon, but they announce themselves differently: symptoms that appear over months rather than years, a deepening voice, marked muscle change or clitoral enlargement. Those features prompt urgent further testing rather than a PCOS label.

The reassuring reality is that for most people, this exclusion round comes back normal and the picture clarifies quickly. Its purpose is not to delay a diagnosis but to make sure the one you receive is the right one.

How is PCOS diagnosed in teenagers?

Adolescence is where the standard criteria become genuinely tricky, because two of the three Rotterdam features are normal parts of growing up. Cycles are commonly irregular for the first couple of years after a first period as the hormonal feedback loops mature. Acne is near-universal. Ultrasound frequently shows ovaries with many follicles in perfectly healthy teenagers, which is why international guidance discourages using imaging to diagnose PCOS in this age group.

Clinicians therefore raise the bar. In a teenager, the diagnosis generally rests on the combination of persistent menstrual irregularity, judged against what is expected for the number of years since the first period, together with clear evidence of excess androgens, whether that is significant hair growth in a male pattern or a testosterone level above the reference range. Acne alone is usually not enough.

When the picture is suggestive but not conclusive, many clinicians describe the young person as “at risk” of PCOS and arrange to reassess once cycles have had more time to settle. This can frustrate families who want an answer, but it protects against labeling a normal developmental phase as a lifelong condition, and against the anxiety that follows.

The practical steps in the meantime look the same regardless of the label: tracking cycles, addressing troubling skin or hair symptoms, and checking blood sugar and weight trajectory where relevant. Cleveland Clinic and the NIH both note that management targets symptoms, so nothing is lost by waiting for diagnostic certainty.

Can you have PCOS without cysts, or cysts without PCOS?

Yes to both, and the two scenarios cause most of the confusion around this condition.

Someone with irregular ovulation and raised androgens meets the criteria regardless of what their ovaries look like. Estimates vary, but a meaningful share of people with a confirmed diagnosis have entirely normal-appearing ovaries on ultrasound. The name is a historical accident from an era when the scan was the most striking finding; if the condition were being named today, it would probably be called something about androgens and ovulation.

The reverse is more common than most people realize. Polycystic-appearing ovaries turn up incidentally in a sizable proportion of scans performed for other reasons, in people with regular cycles, normal hormones and no symptoms. That finding on its own is not PCOS. It is a normal variant, and no treatment or label follows from it. If a radiology report mentions polycystic ovaries and nobody has asked about your periods or checked your hormones, the appropriate next step is a conversation, not a diagnosis.

The distinction also matters for people who have been told they have PCOS on the basis of a scan alone, sometimes years ago. It is reasonable to ask a clinician to revisit that label using the full criteria. Some will have the diagnosis confirmed; others will learn that a single finding was over-interpreted. Either outcome is better than carrying uncertainty.

What tests follow a PCOS diagnosis, and why?

Confirming PCOS is less a finish line than a change of gear. Because the condition is closely linked with insulin resistance, clinicians shift from diagnostic tests to periodic health checks that have nothing to do with the ovaries.

Blood sugar comes first. The CDC reports that more than half of women with PCOS develop type 2 diabetes by age 40, a figure that explains why screening is recommended at diagnosis and then at intervals your clinician sets, rather than once. Depending on your situation this may be a fasting glucose, an HbA1c, or an oral glucose tolerance test, which the NHS and Mayo Clinic describe as more sensitive at picking up early problems with glucose handling.

Cardiovascular risk is assessed alongside: a lipid profile for cholesterol and triglycerides, blood pressure at each visit, and attention to weight trajectory and waist circumference. None of these are unique to PCOS, but the condition shifts the odds enough that clinicians pay closer attention.

Two further checks are easy to overlook. Anxiety and low mood are more common in people with PCOS, and many clinicians now ask about them routinely. Sleep-disordered breathing, including obstructive sleep apnea, is also more frequent, so questions about snoring, waking unrefreshed or daytime sleepiness may come up.

Finally, if periods are very infrequent, the uterine lining is not shed regularly, and a clinician may discuss whether that needs monitoring. This is a conversation to have with your treating team, who will weigh your history and preferences.

Can home hormone kits or period apps diagnose PCOS?

The market for at-home hormone testing has grown quickly, and the appeal is obvious: a finger-prick or saliva sample, a dashboard of numbers, no waiting room. The honest answer is that these products can contribute information but cannot deliver a diagnosis, for three reasons rooted in how the criteria work.

First, PCOS requires exclusion of other conditions, and most consumer kits do not include the exclusion tests or interpret them in context. Second, androgen measurement in women demands sensitive laboratory assays; results near the lower end of the range are technically difficult even for hospital laboratories, and a mail-in sample adds handling variables. Third, the criteria depend on physical examination and cycle history that no kit captures.

Period-tracking apps are a different story. They are genuinely useful for documenting cycle length and frequency, which is exactly the evidence the first Rotterdam criterion rests on. A year of logged dates is more reliable than memory and can shorten the diagnostic conversation considerably. The caveat: some apps predict ovulation from averages and display it as fact. A predicted ovulation day is not evidence that ovulation occurred, and clinicians treat it accordingly.

If you have already used a kit, bring the results; a clinician can decide whether they add anything. What the results should not do is replace the appointment. Knowing where you stand on hormones, blood sugar and heart health is the point of the process, and those answers come from a full assessment.

How long does it take to get a PCOS diagnosis?

In principle, the process is brief: a consultation, a set of blood tests, possibly a scan, and a follow-up visit to put the pieces together. In practice, surveys of people with the condition consistently describe long waits and multiple appointments before the diagnosis is made, and it is worth understanding why.

Some delay is built in and appropriate. Blood tests may need to be timed to a cycle that arrives unpredictably. Someone using hormonal contraception may need a pause before hormones can be interpreted. Teenagers may be deliberately reassessed after a period of watching and waiting. Each of these is the system working as designed, even when it feels slow.

Other delays are avoidable. Symptoms get treated in isolation: acne by one clinician, irregular periods by another, weight by a third, with nobody joining the dots. People are told their periods will settle, or that irregularity is normal, without the follow-up question of how many cycles a year they actually have. And the diagnosis of exclusion sometimes stalls at the exclusion stage, with a normal thyroid result treated as the end of the enquiry.

You can shorten the path. Bring a cycle record. Name all your symptoms in one visit rather than the one that bothers you most. Ask directly whether PCOS is being considered and, if not, why. Clinicians generally welcome a patient who has organized the evidence; it is the same work they would otherwise have to do from scratch.

When should you see a doctor about possible PCOS?

The clearest trigger is a cycle pattern that has become infrequent or unpredictable for more than a few months, particularly if it arrives alongside new or worsening facial or body hair, adult acne that does not respond to ordinary care, or thinning hair at the scalp. Difficulty conceiving after trying for a year, or six months over the age of 35, is another reason to ask about ovulation and PCOS specifically. If a scan done for another reason mentioned polycystic ovaries, that is worth a conversation too, even if you feel well.

Absent periods deserve attention on their own. Missing three or more consecutive cycles when you are not pregnant, breastfeeding or on hormonal contraception should be assessed, because the causes range from PCOS to thyroid disease to other conditions that matter in their own right.

Certain features need prompt rather than routine care. Hair growth or acne that has appeared or accelerated over weeks to months rather than years, a deepening voice, noticeable increase in muscle bulk, enlargement of the clitoris, or breast discharge when you are not breastfeeding can point to rarer hormonal causes that require urgent evaluation. Very heavy bleeding that soaks through protection hourly, bleeding after long gaps that will not stop, or bleeding accompanied by dizziness or faintness warrants same-day contact with a clinician or emergency services.

For everything else, an ordinary appointment with a primary care clinician or gynecologist is the right starting point. Arrive with your cycle dates, a list of symptoms, and your questions written down. The diagnosis may take more than one visit, but it starts with that one.

Frequently asked questions

Can PCOS be diagnosed with a blood test alone?

No. Blood tests can show raised androgens and rule out thyroid, pituitary and adrenal causes, but the diagnosis also requires either a history of irregular ovulation or a characteristic ultrasound appearance. A single raised testosterone result with regular cycles and normal ovaries does not meet the criteria, and a normal result does not exclude PCOS in someone with irregular cycles and clear physical signs.

Do I need an ultrasound to be diagnosed with PCOS?

Not necessarily. Under the Rotterdam criteria, irregular or absent ovulation combined with clinical or biochemical evidence of excess androgens is sufficient, and both the NHS and Mayo Clinic note that ultrasound can be skipped in that situation. A scan is used when only one criterion is present, when another cause of symptoms needs checking, or when the uterine lining should be assessed after long gaps between periods.

What day of my cycle should PCOS blood tests be done?

Ideally in the first few days after bleeding starts, when LH, FSH and estrogen are at a stable baseline that laboratories can interpret. Testosterone is usually highest in the morning, so early appointments are often preferred. If your cycles are too irregular to time, a random sample is acceptable and the results are simply interpreted with that in mind.

Will hormonal contraception affect PCOS test results?

Yes, substantially. Combined hormonal contraception raises the protein that binds testosterone and suppresses ovarian androgen production, so blood tests can look normal, and the regular withdrawal bleeds hide your natural cycle pattern. Your clinician may recommend a pause before testing or may rely on your pre-contraception history and physical signs; that decision is made together with the prescribing clinician.

Can you have PCOS with regular periods?

Yes. A person with regular cycles can meet the criteria if they have evidence of excess androgens, such as male-pattern hair growth or a raised testosterone level, together with polycystic-appearing ovaries on ultrasound. This phenotype is sometimes missed because clinicians and patients assume irregular periods are essential. Occasionally, cycles that appear regular are not actually ovulatory, which a mid-cycle blood test can check.

What other conditions are ruled out before diagnosing PCOS?

Thyroid disorders, raised prolactin and non-classic congenital adrenal hyperplasia are routinely screened for with blood tests, because each can cause irregular periods or androgen-related symptoms. Rarer possibilities such as Cushing’s syndrome or androgen-secreting tumors are considered when symptoms are unusually rapid or severe. PCOS is formally a diagnosis of exclusion, so these checks are a required step rather than an optional extra.

Is AMH used to diagnose PCOS?

Not as a standalone criterion at present. Anti-Müllerian hormone is often elevated in PCOS because it is produced by the many small follicles, and researchers are studying it as a possible alternative to ultrasound. International guidance treats it as supportive information rather than a diagnostic test, partly because assays and cutoffs are not yet standardized. Your clinician may include it but will interpret it alongside everything else.

Why is PCOS harder to diagnose in teenagers?

Because irregular cycles, acne and ovaries with many follicles are all common and normal in the years after a first period. Guidance therefore discourages using ultrasound in adolescents and sets a higher bar: persistent irregularity beyond what is expected for their stage, plus clear evidence of excess androgens. Many clinicians describe a teenager as at risk and reassess later rather than applying a lifelong label prematurely.

What happens after a PCOS diagnosis is confirmed?

Attention shifts to long-term health checks. Blood sugar is tested at diagnosis and then at intervals, because the CDC reports that more than half of women with PCOS develop type 2 diabetes by age 40. Cholesterol, blood pressure and weight trajectory are monitored, and clinicians increasingly ask about mood and sleep. Management is tailored to your symptoms and goals in discussion with your treating team.

Can an at-home hormone test diagnose PCOS?

No. Consumer kits may report hormone values, but the diagnosis depends on excluding other conditions, on physical examination and on cycle history, none of which a kit provides. Androgen measurement in women is also technically demanding even for hospital laboratories. Bring any results you have to an appointment, where a clinician can decide whether they add useful information to a full assessment.

References

• NHS: Polycystic ovary syndrome: Diagnosis
• MedlinePlus: Polycystic Ovary Syndrome
• NIH (NICHD): How do health care providers diagnose PCOS?
• Cleveland Clinic: Polycystic Ovary Syndrome (PCOS)

This article is for general information only and is not a substitute for professional medical advice. Please consult a qualified doctor about your individual situation.