You have a printout, or a portal screen, with a column of numbers and possibly an arrow beside one of them. You typed PCOS hormone levels chart into a search box, because what you want is a table to hold your results against so you can see where you stand.
That table is the thing this page is going to argue does not exist. Not because it is being withheld, and not because the information is complicated — because of something specific about how these particular substances are measured. Understanding it takes about five minutes and it changes what you do with the sheet in your hand.
The same blood, two different numbers
The 2023 international guideline asks laboratories to measure total testosterone using tandem mass spectrometry — written LC-MS/MS — rather than direct immunoassay, on the grounds of accuracy. Free testosterone it asks to be assessed by calculation, equilibrium dialysis or ammonium sulfate precipitation.
That is a recommendation about method, and it exists because method changes the answer. Female testosterone concentrations are low, and direct immunoassays were built for the much higher concentrations found in men; at the low end they are imprecise in ways that matter. Two laboratories running the same sample by different methods do not produce the same figure.
Which means a chart of “normal PCOS hormone levels” found on the internet has a problem it cannot solve: it does not know which method produced your number. It is comparing your result against a range built from someone else’s instrument. The arithmetic looks fine and the conclusion is unfounded.
The chart you need is already on your report
Beside your result, your laboratory prints its own reference range. That range was built for the assay that machine runs, which is the assay that produced your number. It is the only range that is actually about your result.
This is why a report can say “normal” while an internet table says “high”, and why the internet table is the one that is wrong. It is also why comparing your figure to a friend’s, or to one quoted in a forum, tells you close to nothing unless you happen to know both laboratories used the same method.
If you want one question to take to an appointment from this whole page, it is this: which method did the laboratory use? It is a fair question, the answer is on file, and it determines how much weight the number can carry.
What each test is actually for
| Test | What it is | What the guideline says |
|---|---|---|
| Total testosterone | All circulating testosterone, bound and unbound | First-line for biochemical hyperandrogenism; should be measured by LC-MS/MS rather than direct immunoassay |
| Free testosterone | The unbound fraction | First-line alongside total; by calculation, equilibrium dialysis or ammonium sulfate precipitation |
| Free androgen index | A calculation from testosterone and SHBG | The common form of calculated free testosterone; moves whenever SHBG moves |
| Androstenedione, DHEAS | Other androgens, partly adrenal | Considered where testosterone is not elevated, noting poorer specificity and DHEAS falling with age |
| AMH | A marker of follicle numbers | May be used instead of ultrasound for the third criterion in adults; not in adolescents |
| Thyroid, prolactin and others | Not PCOS markers | Ordered to exclude conditions that look like PCOS, not to confirm it |
Read down the right-hand column and a pattern shows up. Almost every entry carries a condition — a method, an age group, a caveat about specificity. None of them carries a number. That is not an omission on this page; it is what the guideline itself does.
No single number diagnoses PCOS anyway
Even a testosterone measured by the right method on the right machine does not, by itself, answer the question you are asking the chart.
PCOS is diagnosed against the Rotterdam criteria — two of three: irregular or absent ovulation, hyperandrogenism, and polycystic ovarian morphology on ultrasound or, in adults, a raised AMH. One blood result is a component of one of those three.
And that criterion has two routes into it. Hyperandrogenism can be met clinically — hirsutism, acne, hair loss — as well as biochemically. Which is why bloods inside the reference range do not close the question for someone with the clinical picture, and why a diagnosis is not something a laboratory hands out.
What most charts get wrong
The charts circulating online tend to make three moves, and each one is a small step away from what the evidence supports.
They publish a single range with no method attached. Which is the whole problem above, presented as though it were solved.
They assign meaning to ratios. The LH to FSH ratio is the usual example — widely reproduced, and not among the first-line assessments the guideline names for biochemical hyperandrogenism. A number can be real, reproducible, and still not be a criterion.
They treat the result as the verdict. A figure against a range becomes a pass or a fail, and then a diagnosis, in about two lines. Rotterdam needs two of three findings and a clinician who has excluded the conditions that mimic PCOS. No table does that.
What to do with the sheet in front of you
Three things are worth doing, and none of them involves a chart.
Read the range printed beside each result, not one from elsewhere. It is the one built for the machine that produced your number.
Ask which method was used, particularly for testosterone. If it was a direct immunoassay, the guideline’s preference is on record and it is reasonable to ask whether a mass spectrometry measurement is available.
Keep the results. A single value is a point; the same test a year later is a direction, and direction is what an appointment can actually work with. That is true of every number on the sheet, and it is the one thing you can do with a result today that makes it more useful later.