AMH Age Percentile Interpreter

AMH falls throughout reproductive life, so a single reference interval is close to useless — 1.9 ng/mL is low at 25 and unremarkable at 37. This reads a result against a named age-stratified nomogram, and refuses to give a centile when the assay is not the one the nomogram was built on.

Last updated 17 Sep 2026

AMH against age

Value + age + assay → centile

AMH resultEnter the number exactly as the report prints it, then set the unit below. AMH is reported in ng/mL (identical to µg/L) in most of the world and in pmol/L across much of Europe.

Unitpmol/L is divided by 7.14286 to reach ng/mL, from a molecular weight of 140,000 for the AMH dimer: 1 ng/mL = 1×10⁻⁶ g/L, and 1×10⁻⁶ ÷ 140,000 = 7.14 pmol/L. That factor is exact arithmetic; the assay calibration underneath it is not.

Age (years)The nomogram used here covers 20 to under 55 years. Outside that range this page will say so rather than extrapolate — AMH rises through childhood to a peak in the early twenties, and the shape of that limb is not described by an adult nomogram.

Assay the result came fromThis is not a formality. The nomogram below was built on the Beckman Coulter Gen II assay. Elecsys reads lower than Gen II and picoAMH, and about half of measurements cannot be reliably converted between platforms, so a centile from one assay does not transfer to another.

What the result is being read forThe two questions have different answers. For ovarian reserve a centile against age is the right frame. For PCOS it is not: AMH there is a yes or no against the laboratory's own cut-off, and no guideline gives a number.

Between the 5th centile and the median for ageExample

AMH 1.9 ng/mL, age 36, Beckman Coulter Gen II, read for ovarian reserve

How the reading is made

pmol/L ÷ 7.14286 = ng/mL · then compared against the 5th centile, median and 95th centile for the age band
Nomogram: Du 2016, Beckman Coulter Gen II, 1,590 healthy Chinese Han women aged 20–55

7.14286: pmol/L per ng/mL, from a molecular weight of 140,000 for the AMH dimer. 1 ng/mL = 1×10⁻⁶ g/L; 1×10⁻⁶ ÷ 140,000 = 7.1429×10⁻¹² mol/L = 7.14 pmol/L. The arithmetic is exact
age bands: 20 to under 25, 25 to under 30, 30 to under 33, 33 to under 37, 37 to under 40, and 40 to under 55 years — the bands the nomogram publishes, not bands chosen here
the population: healthy Chinese Han women, 1,590 of 2,055 recruited, nationwide. That is the population the centiles describe. AMH distributions differ between populations, so a centile read for a woman outside it is indicative rather than exact, and this is stated here rather than buried
the assay: Beckman Coulter Gen II. A result from Elecsys or picoAMH is a different measurement; the page declines to give a centile for those rather than pretending the number transfers
why age, not one interval: AMH peaks around 20 to 25 years and declines quadratically thereafter. A single all-ages reference interval is wide enough to call almost nothing abnormal in a young woman and almost everything abnormal in an older one

Worked example

AMH 1.9 ng/mL, age 36, Beckman Coulter Gen II, read for ovarian reserve
Age 36 falls in the nomogram's 33 to under 37 band
That band's centiles are 5th 0.87, median 3.74, 95th 9.76 ng/mL
1.9 is above 0.87 and below 3.74
So the result is within the reference range for age, below the median
Read against a single all-ages interval, or against a 25-year-old's values, 1.9 ng/mL would look low — it is not
Expect a mid-range response to stimulation. The result says nothing about natural conception

Du 2016 age-specific centiles — Beckman Coulter Gen II, healthy Chinese Han women

Du X, Ding T, Zhang H, et al. *Reprod Sci* 2016;23(8):1019–1027. 1,590 healthy women of 2,055 recruited, nationwide, Beckman Coulter Gen II assay, AMH declining with age on a quadratic model. The pmol/L column is the median multiplied by 7.14286 and is arithmetic, not a second measurement. The site's AMH unit converter carries a separate set of age-banded intervals measured on the Roche Elecsys platform; the two are not interchangeable.
Age (years)	5th centile (ng/mL)	Median (ng/mL)	95th centile (ng/mL)	Median in pmol/L
20 to under 25	2.06	6.23	12.66	44.5
25 to under 30	1.77	5.65	13.83	40.4
30 to under 33	1.48	4.55	11.45	32.5
33 to under 37	0.87	3.74	9.76	26.7
37 to under 40	0.56	2.78	9.49	19.9
40 to under 55	0.08	1.09	5.70	7.8

What AMH does and does not predict

Practice Committee of the ASRM, *Fertil Steril* 2020;114(6):1151–1157. The two rows that say yes are the reason AMH is measured; every row that says no is a use it is put to anyway.
Question	Does AMH answer it	Source
Oocyte yield at stimulation	Yes — well demonstrated, together with antral follicle count	ASRM committee opinion 2020
Poor or excessive response to gonadotrophins	Yes — this is the established clinical use	ASRM committee opinion 2020
Fecundability, or time to natural pregnancy	No — markers of ovarian reserve were poor predictors	ASRM committee opinion 2020
Cumulative probability of pregnancy	No — women with low AMH had similar cumulative pregnancy rates	ASRM committee opinion 2020
Oocyte quality	No — AMH estimates pool size, not competence	ASRM committee opinion 2020
Whether to offer or withhold IVF	No — extremely low values should not be used to refuse treatment	ASRM committee opinion 2020
Whether to freeze eggs	No — the decision rests on reproductive plans and age	ASRM committee opinion 2020
Whether contraception is still needed	No — AMH is not a contraceptive indicator	Follows from the fecundability findings above

A hormone with no single normal range

Anti-Müllerian hormone is secreted by the granulosa cells of small growing follicles, which makes it a reasonable proxy for how many follicles are left. It also means it falls, continuously, from a population peak in the early twenties to near-undetectable at the menopause. A laboratory that reports AMH against one interval spanning all reproductive ages is therefore reporting against an interval so wide that a young woman has to be profoundly depleted to fall outside it, and an older woman has to be polycystic to stay inside it. The only reading that carries information is the one against age — which is why this page asks for a date of birth before it will say anything.

The nomogram used here is named on the page rather than implied: Du and colleagues measured AMH in 1,590 healthy Chinese Han women aged 20 to 55, nationwide, on the Beckman Coulter Gen II assay, and published fifth centiles, medians and ninety-fifth centiles for six age bands. Naming it is not pedantry, because two things about it constrain what a centile from it means. The population is one population, and AMH distributions differ between populations, so the reading is indicative outside it. And the assay is one assay. AMH immunoassays are not standardised against a common reference material: a direct comparison of the Gen II, picoAMH and Elecsys platforms found Elecsys measuring lower than both of the others, with correlation weakest in the middle of the range and around half of measurements not reliably convertible between platforms. That is why this page refuses to print a centile when the assay selector says anything other than Gen II. A number that looks authoritative and is not transferable is worse than no number.

The practical consequence of assay drift is bigger than the centile question. If a woman's AMH was 3.1 last year and 2.2 this year, the first thing to establish is whether both were measured in the same laboratory on the same platform, because a change of either can account for the whole difference. Several other things move AMH without the ovary changing: combined oral contraception suppresses it, ovarian surgery reduces it, a higher body mass index is associated with lower values, and cycle day contributes a little. Those are described in more detail on the AMH unit converter, which also handles the unit arithmetic if you only need to move between ng/mL and pmol/L.

What remains is the hardest part to say clearly, because it is the opposite of how the test is marketed. AMH predicts two things well: oocyte yield at stimulation, and whether the response to gonadotrophins will be poor or excessive. That is a genuinely valuable pair of predictions and it is why the assay exists in reproductive medicine. It does not predict natural fertility. The ASRM committee opinion reviewed the evidence and concluded that markers of ovarian reserve were poor predictors of fecundability, of cumulative probability of pregnancy, and of the incidence of infertility, and that women with low AMH had similar cumulative pregnancy rates to those with normal values. It does not measure egg quality — pool size and oocyte competence are different properties. It should not be used to refuse IVF, and ASRM says so in terms. It should not be used to push anyone towards egg freezing. And it is not a contraceptive indicator in either direction: a low AMH is not contraception, and a high one is not a reason to expect conception. If the question is whether AMH raises the possibility of polycystic ovary syndrome, that is a different question with a different answer — see the Rotterdam criteria, where AMH may substitute for an ultrasound but never joins it, and where no guideline supplies a number.

Frequently asked questions

What is a normal AMH for my age?

On the Du 2016 nomogram, measured on the Beckman Coulter Gen II assay in healthy women, the median is 6.23 ng/mL at 20 to under 25 years, 5.65 at 25 to under 30, 4.55 at 30 to under 33, 3.74 at 33 to under 37, 2.78 at 37 to under 40 and 1.09 from 40 to under 55. The fifth centiles are much lower — 0.87 ng/mL at 33 to 37, for example — so a value that looks low may be entirely normal for age.

How do I convert AMH from pmol/L to ng/mL?

Divide by 7.14286. The factor comes from the molecular weight of the AMH dimer, 140,000: one ng/mL is 1×10⁻⁶ g/L, and dividing by 140,000 gives 7.14×10⁻¹² mol/L, which is 7.14 pmol/L. So 25 pmol/L is 3.5 ng/mL. The arithmetic is exact; the assay calibration underneath it is not, which is why results do not transfer between platforms.

Does a low AMH mean I cannot get pregnant?

No. AMH estimates the size of the remaining follicle pool and predicts response to ovarian stimulation. The ASRM committee opinion found markers of ovarian reserve to be poor predictors of fecundability, cumulative probability of pregnancy and incidence of infertility, with low-AMH women showing similar cumulative pregnancy rates to those with normal values. ASRM also states that extremely low values should not be used to refuse IVF.

Why is my new AMH different from my last one?

Check the laboratory and the platform first. Elecsys reads lower than Gen II and picoAMH, correlation between assays is weakest in the middle of the range, and about half of measurements cannot be reliably converted between platforms — so a change of assay can account for the whole difference. Hormonal contraception, ovarian surgery, a higher BMI and cycle day all move the result as well.

Can AMH tell me when I will reach the menopause?

Not with useful precision for an individual. AMH declines towards the menopause and correlates with it at population level, but the prediction interval around any one woman's date is wide. It is also not a contraceptive indicator: a low AMH does not mean contraception can be stopped, and a normal one does not mean conception is likely.

Is there an AMH level that diagnoses PCOS?

No guideline gives one. The 2023 PCOS guideline permits AMH for defining polycystic ovarian morphology in adults but supplies no threshold, instructing laboratories to use population- and assay-specific cut-offs instead. It also says AMH must not be used as a single test for the diagnosis, must not be used in adolescents, and must not be combined with an ultrasound — either one or the other, to limit over-diagnosis.

Related calculators

References

Du X, Ding T, Zhang H, et al. Age-Specific Normal Reference Range for Serum Anti-Müllerian Hormone in Healthy Chinese Han Women: A nationwide Population-Based Study. Reprod Sci. 2016;23(8):1019–1027. doi:10.1177/1933719115625843 — Beckman Coulter Gen II assay, 1,590 healthy women aged 20–55; 5th centile, median and 95th centile by age band.
Barbagallo F, van der Ham K, Willemsen SP, Louwers YV, Laven JSE. Age-related Curves of AMH Using the Gen II, the picoAMH, and the Elecsys Assays in Women With Polycystic Ovary Syndrome. J Clin Endocrinol Metab. 2024;109(10):2561–2570. doi:10.1210/clinem/dgae153 — "lower AMH values were measured by the Elecsys AMH assay than both the Gen II AMH and the picoAMH assays"; roughly half of measurements not convertible between assays; age-percentile distributions comparable only with assay-specific cut-offs.
Practice Committee of the American Society for Reproductive Medicine. Testing and interpreting measures of ovarian reserve: a committee opinion. Fertil Steril. 2020;114(6):1151–1157. doi:10.1016/j.fertnstert.2020.09.134 — AMH and AFC predict oocyte yield and poor or excessive stimulation response; markers of ovarian reserve are poor predictors of fecundability and cumulative pregnancy; extremely low values should not be used to refuse IVF; ovarian reserve markers should not be used to promote planned oocyte cryopreservation.
Steiner AZ, Pritchard D, Stanczyk FZ, et al. Association Between Biomarkers of Ovarian Reserve and Infertility Among Older Women of Reproductive Age. JAMA. 2017;318(14):1367–1376. doi:10.1001/jama.2017.14588
Teede HJ, Tay CT, Laven JJE, et al. Recommendations From the 2023 International Evidence-based Guideline for the Assessment and Management of Polycystic Ovary Syndrome. J Clin Endocrinol Metab. 2023;108(10):2447–2469. doi:10.1210/clinem/dgad463 — recommendations 1.5.1 to 1.5.7 on AMH, including that no numeric threshold is given and that laboratories should use population- and assay-specific cut-offs.

Medical Disclaimer: The tools and content provided here are for educational and reference purposes only. They are not intended to substitute for professional medical advice, diagnosis, or treatment. Clinical decisions should always be based on the comprehensive assessment of a qualified healthcare professional.