PCOS With Normal Labs: Why the Bloods Can Look Fine

Can PCOS blood tests really come back completely normal?

Yes — roughly 19% of people with PCOS in unselected (non-referral) populations meet the diagnosis through irregular ovulation plus polycystic ovaries on ultrasound alone, with every hormone on a standard androgen panel sitting inside its printed reference range, according to a 2016 meta-analysis pooling nearly 13,800 diagnosed patients. The 2023 international evidence-based guideline still uses the Rotterdam criteria first published in 2003: two of three findings — irregular or absent ovulation, clinical or biochemical signs of excess androgens, and polycystic ovaries on ultrasound (or an elevated AMH in adults) — confirm the diagnosis. Nothing in that rule requires every leg to be abnormal. A completely clean blood panel closes off exactly one of three possible routes to the diagnosis, not all three — the other two, and how they combine into a phenotype, are covered in the full work-up.

Note: in May 2026, PCOS was renamed polyendocrine metabolic ovarian syndrome, or PMOS, by a global consensus of more than 50 organisations. Nothing about which labs count as normal changed — only the name on the chart did. This article uses PCOS, since that is still what most readers type into a search bar.

*Pooled estimates from a 2016 meta-analysis of 41 studies; figures for referral (specialist clinic) populations skew higher toward phenotype A and lower toward phenotype D, since more severe presentations are more likely to be referred in the first place.

Can PCOS be misdiagnosed?

Yes, in both directions, and the guideline itself names this as a known problem worth correcting for. Rotterdam criteria applied loosely — polycystic-appearing ovaries treated as sufficient on their own, or irregular periods in the two years after a first period treated as pathological rather than a normal part of reproductive maturation — over-diagnose PCOS in people who do not have it. Under-diagnosis runs the other way: a normal-looking androgen panel gets treated as a full rule-out when it has only excluded one of three legs of the criteria, and no one goes on to properly assess cycle history or order an ultrasound or AMH. Both mistakes trace back to the same habit — reading one test result as the whole diagnostic picture instead of one input into a three-part rule. Diagnostic delay compounds the problem: irregular cycles are easy to attribute to stress, recent contraceptive use, or “just how your body is” for years before anyone orders the panel that would actually test a Rotterdam leg, and by the time labs are drawn a person may have already been told, incorrectly, that normal-looking skin and hair meant PCOS was off the table.

Can you have PCOS without hirsutism?

Yes — clinical hyperandrogenism is assessed by visible signs like hirsutism, but ethnic background changes how much excess androgen it takes to produce a visible pattern, which is exactly why the Rotterdam criteria accept either clinical or biochemical hyperandrogenism rather than requiring visible hair growth specifically. Someone with a laboratory-confirmed elevated free androgen index and no visible hirsutism still satisfies that leg of the criteria on the blood result alone. And, independent of ethnicity, phenotype D — the roughly one-in-five pattern in the table above — requires no hyperandrogenism, clinical or biochemical, at all. Absence of hirsutism on examination is therefore never, by itself, evidence against a PCOS diagnosis.

Can you have PCOS without insulin resistance?

Yes — insulin resistance is common in PCOS but it is not part of the Rotterdam criteria, and a meaningful subset of people who are lean and diagnosed with PCOS test with normal insulin sensitivity on formal testing. The androgen elevation in that group tends to run through pituitary LH output rather than through the insulin-driven pathway that dominates in heavier or insulin-resistant phenotypes. A fasting glucose, HbA1c, or even an oral glucose tolerance test coming back inside range does not argue against the diagnosis; it simply says this particular metabolic mechanism is not the one driving your presentation.

Functional ranges vs. lab reference ranges: what’s the difference, and which one matters?

The reference range printed on your lab report is built from a large population of people without the condition being tested for, generally set to capture the central 95% of that population’s results — which is also why a result sitting just outside range is common in healthy people and does not automatically signal disease. “Functional” or “optimal” ranges promoted outside of clinical laboratories are a different thing entirely: narrower bands with no population study behind them and no demonstrated link to a diagnostic outcome. A testosterone or DHEA-S result that a functional-medicine source calls “high normal” or “suboptimal” is being compared to a number nobody validated against actual health outcomes. The range that determines whether your bloodwork satisfies the biochemical-hyperandrogenism leg of Rotterdam is the one printed on your own report, from your own lab, using your lab’s own assay — not a number from an article, a supplement company, or a wellness account. A testosterone or DHEA-S result a functional-medicine chart flags as “suboptimal” while your lab reports it as within range has not been shown to predict anything a validated reference range doesn’t already capture — treat that distinction as marketing language layered on top of a real number, not a second, more accurate number.

What to actually do if your labs are normal but symptoms continue

The Rotterdam criteria require confirming two of three legs, not one, so the next step after a normal androgen panel is checking the other two as part of a full PCOS diagnostic work-up — not repeating the same panel and hoping for a different result. That means a documented ovulation pattern (cycle length tracked for at least two to three months, not recalled from memory) and either a transvaginal ultrasound or an AMH level, per the 2023 international guideline. Two circumstances are worth ruling out before concluding a panel is genuinely uninformative. First, hormonal contraception suppresses LH, FSH, and ovarian androgen output while raising SHBG, so bloodwork drawn on the combined pill reads artificially reassuring across every androgen measure Rotterdam cares about; the guideline recommends retesting after stopping, under a clinician’s guidance, rather than treating an on-pill panel as definitive. Second, a borderline total testosterone run on a standard immunoassay is a candidate for a repeat draw using liquid chromatography-tandem mass spectrometry, which is materially more accurate at the low concentrations typical in women — a genuinely normal result on that more precise method carries more weight than a borderline one on the assay most primary-care panels default to.

Who this doesn’t help

Reading this article will not tell you whether your specific results are normal — that requires comparing your number to the range printed on your own report, not a number quoted here. It also will not resolve a genuinely ambiguous case: someone with mildly irregular cycles, borderline ultrasound findings, and androgens at the top of the reference range needs a clinician weighing the whole picture, not a rule that can be read off a table. And normal labs do not rule out related conditions that mimic PCOS — thyroid disease, high prolactin, and non-classic congenital adrenal hyperplasia all deserve their own tests regardless of how the androgen panel reads, because a normal PCOS panel says nothing about any of those three.