AMH levels by age: reference values chart in ng/ml

Anti-Müllerian hormone (AMH) in the blood reflects how many follicles are still present in the ovaries. Over the years, the level falls until menopause (Nelson, Human Reproduction Update 2023). Here you will find anti-Müllerian hormone levels by age from the reference study for the Roche lab test, in ng/ml and pmol/l, plus a comparison table for a second widely used lab test.

These numbers describe groups of women. They do not classify your own test result, and they are no substitute for a conversation with your doctor.

AMH levels by age: reference values in a table

The values come from the manufacturer, Roche, for the Elecsys AMH Plus lab test (method sheet V2.0, as of February 2024). They are based on study RD001727 of 887 apparently healthy women of Caucasian descent who were not using contraceptives. The women were 20 to 50 years old and had regular cycles, and their blood samples were measured in 5 European labs (Anckaert 2016). The values in ng/ml and in pmol/l both appear this way in the method sheet; nothing was converted.

• P50 (median): Half of the women in this age group were below this value, and the other half were above it.
• P5 to P95: The middle 90% of the women fell in this range.
• P2.5 to P97.5: The middle 95% of the women fell in this range.

At the low end, the test reaches its measuring limits. At ages 40 to 44, P2.5 is below the limit of quantitation of 0.03 ng/ml, so the table shows “< 0.03” there. At ages 45 to 50, P2.5 and P5 are at the limit of detection of 0.01 ng/ml, which is the limit of what the test can detect at all. The table therefore shows “< 0.01” there. In the pmol/l view, these places show “< 0.21” (limit of quantitation, ages 40 to 44) and “< 0.07” (limit of detection, ages 45 to 50).

The age selector only highlights the matching row. The page does not classify your own lab value.

How AMH changes with age

The table compares different women of different ages; it does not show how AMH changes in individual women over time. The median is 4.00 ng/ml at ages 20 to 24 and 2.81 ng/ml at ages 30 to 34. In the 40 to 44 group, at 0.882 ng/ml, it is less than half as high as in the 35 to 39 group, at 2.00 ng/ml. At ages 45 to 50, it is 0.194 ng/ml.

In a large US analysis of fertility patients, measured with an older lab test, the median fell by about 0.2 ng/ml per year of age until age 35, and by about 0.1 ng/ml afterward (Seifer 2011). These steps cannot be applied directly to the Roche values above.

The ranges overlap a lot. The top 5% of women aged 40 to 44 were above 4.44 ng/ml, which is above the median of women aged 20 to 24. In a model based on data from healthy women, AMH peaked on average at about age 24.5, and age explained only 34% of the differences between women (Kelsey 2011).

How to read the table

A percentile is the value below which a certain share of the study group fell. Below P5 were 5 out of 100 women in the age group; below P95 were 95 out of 100. This is not a diagnosis, and it says nothing about how easily an individual woman will get pregnant.

Labs adopt the Roche table in two variants, as their test directories show: Ulm University Hospital and MVZ Medizin-Zentrum Dortmund use P2.5 to P97.5, and Endokrinologikum Hamburg uses P5 to P95. Ulm rounds, for example, to 0.58, 0.15, and 0.03 ng/ml. That is why the table above shows both variants. Other labs use other tests or their own limits.

Roche itself states that each lab should check whether the values can be applied to its patients. For your own result, the reference range of your lab applies, and your doctor interprets it.

The ends of the range are less certain than the middle because each age group includes only 138 to 169 women. P2.5 for women aged 35 to 39 is 0.147 ng/ml, and the 95% confidence interval ranges from 0.053 to 0.474 ng/ml. P5 and P95 are more robust.

ng/ml or pmol/l? Converting between them

Lab reports give AMH in ng/ml or in pmol/l. Some labs write µg/l, which is numerically the same as ng/ml.

• ng/ml × 7.14 = pmol/l
• pmol/l × 0.14 = ng/ml

Roche, the British Association for Clinical Biochemistry, and Ulm University Hospital all give the same factor. It comes from the molar mass of the AMH molecule, 140 kilodaltons (Gassner & Jung 2014). An example: 4.00 ng/ml times 7.14 equals 28.56 pmol/l; the method sheet gives 28.6 pmol/l, rounded.

Why AMH tables online differ

The main reason is the lab test. There is no international standard for AMH, only a WHO reference reagent (16/190). The Roche Elecsys test measures 24% to 28% lower than the older, modified Gen II ELISA, as a comparison of 1,729 routine samples showed (Anckaert 2016). The two widely used automated tests also differ from each other: in 293 blood samples, Elecsys values were about 0.81 times the Access 2 values (Han 2022).

The reference group matters as well. Many older tables are based on fertility patients and older testing methods. Seifer 2011 analyzed 17,120 women from fertility clinics in the US, Nelson 2011 analyzed 9,601 fertility patients, and Almog 2011 looked at 3,871 women with fertility problems, where the test kit itself affected the value. We could not check the tables from these studies against the full text, so we do not show them.

Reference groups also differ for tests from the same manufacturer. In a Korean population study using the Elecsys AMH test (not AMH Plus), the medians were in the same range as Roche’s, but the ends of the range differed clearly: P2.5 at ages 30 to 34 was 1.14 pmol/l there, compared with 4.11 pmol/l in the Roche reference study. The Korean study gives no information about use of the pill, polycystic ovary syndrome (PCOS), or cycle.

Besides Elecsys, the Access test from Beckman Coulter is widely used. Its reference is based on 666 apparently healthy women aged 18 to 45 from 5 centers (FDA market clearance 510(k) K170524). It gives only the median and the P2.5 to P97.5 range, and the age groups are divided differently.

Source: FDA 510(k) Decision Summary K170524 (2017), Expected values, Serum. Reference group: 666 apparently healthy women aged 18 to 45 years at 5 sites. Further percentiles are not given there.

Your lab can tell you which test was used for your result.

What the AMH level tells you and what it does not

AMH gives a rough idea of how many eggs can be retrieved with hormone stimulation, for example in IVF or when freezing eggs (ESHRE 2024). To predict a poor or strong response of the ovaries to stimulation, ESHRE and NICE recommend AMH or the number of antral follicles on ultrasound (ESHRE 2025, NICE 2026).

AMH does not reliably reflect egg quality (ESHRE 2024), and the level does not predict natural pregnancy. A US study of 750 women aged 30 to 44 with no known fertility problems, who had been trying to get pregnant for no more than 3 months, shows this (Steiner 2017):

The differences were not statistically significant. A positive test was counted, not a birth, and the group with AMH below 0.7 ng/ml was small. The professional societies reach similar conclusions:

• ASRM (USA, 2020): In women without a fertility problem, AMH is not a fertility test. The value should be used neither to advertise elective egg freezing nor to deny treatment. Age is more predictive.
• ACOG (USA, 2019): In women without a diagnosis of infertility, a single AMH value does not predict time to pregnancy. It is also not recommended for predicting miscarriage.
• ESHRE (2023) and NICE (2026): AMH should not be used to predict spontaneous pregnancy. According to ESHRE, women with regular cycles do not need an ovarian reserve test for this.
• DGGEF, DGRM, and BVF (2014): They reject general screening and a self-initiated fertility check. Without difficulty conceiving, AMH is either not informative or informative only to a very limited extent.

AMH also predicts the chance of a live birth after IVF only imprecisely (Iliodromiti 2014, ESHRE 2025). For women under 40, the level estimates the timing of menopause only with an uncertainty of 2 to 12 years (Nelson, Human Reproduction Update 2023). What matters when freezing eggs is explained in our elective egg freezing guide.

AMH and IVF: data from the German IVF Registry

In its 2024 yearbook, the German IVF Registry (Deutsches IVF-Register, D.I.R.) analyzed results by AMH for the first time. The analysis is based on 42,572 fresh cycles with IVF or ICSI from 2024 with documented AMH. According to the registry, that is about 70% of these treatment cycles. AMH has only been a required field there since December 2023. The registry forms three groups: below 1.0 ng/ml, 1.0 to 3.5 ng/ml, and above 3.5 ng/ml.

Source: D.I.R. Yearbook 2024, pp. 12 and 13 (cycle year 2024, 42,572 fresh IVF/ICSI cycles with an AMH value). The figures count treatment cycles of women receiving fertility treatment in Germany, not women in the general population. The shares are calculated from the case numbers. The registry itself describes the grouping as rough and points to differences between the lab tests.

The share shows how the cycles of an age group are distributed across the AMH groups. Under age 35, 16.5% of cycles were below 1.0 ng/ml, and from age 40 on it was 51.6%. With higher AMH, more eggs were retrieved on average: across all age groups, 4.62 at below 1.0 ng/ml, 9.33 at 1.0 to 3.5 ng/ml, and 12.38 at above 3.5 ng/ml.

The clinical pregnancy rate per fresh transfer differed more between the age groups than between the AMH groups. For women under 35 with AMH below 1.0 ng/ml, it averaged 33.3%, and for women aged 40 and older with AMH above 3.5 ng/ml, 22.3%.

These notes can help you read the data:

• The counts are cycles of women in fertility treatment, not individual women.
• The rate applies per fresh transfer with a known outcome and counts pregnancies, not births. With AMH below 1.0 ng/ml, 78.8% of cycles across all age groups reached transfer, which gave 18.7% clinical pregnancies per started cycle.
• The registry itself calls its grouping “very rough” and points out that different tests can lead to different AMH values.

Our IVF Calculator with AMH also shows registry data by age. The pages on IVF and ICSI explain the process.

How many eggs does a woman have?

According to the US professional society ACOG, a woman has 1 to 2 million eggs in both ovaries at birth and 300,000 to 500,000 at puberty. At age 37, there are about 25,000, and at age 51, the average age at menopause in the US, about 1,000 (ACOG 2025). The supply is established before birth and only declines afterward (Wallace & Kelsey 2010). Which number you find for a given age depends on the counting method. Wallace & Kelsey model the resting follicles per ovary from 325 histological data points, while ACOG Committee Statement No. 22 from 2025 gives numbers for both ovaries from older counts. The columns are therefore not directly comparable.

ASRM gives about 500,000 to 1 million eggs at birth (ASRM 2020). The differences between women are enormous: at age 35, the model’s 95% prediction interval ranges from 1,900 to 135,000 follicles per ovary. AMH gives only a hint about this number. What it means for getting pregnant after 40 is covered in the article Getting pregnant at 40.

Can the AMH level improve or go up again?

A second measurement can come out higher without anything having changed in the ovarian reserve. In the same woman, AMH varies by 20% on average over the course of a cycle, mostly for biological reasons. In 29% of the women, a value from a different cycle day would have led to a different AMH category in the study (Hadlow 2016, 38 women).

After stopping the pill, AMH rose by 53% on average in 68 women who had taken it for many years, and it reached a plateau after about 2 months (Landersoe, Reproductive BioMedicine Online 2020). In the Dutch Doetinchem study, AMH was lower with current pill use and current smoking, but not with past use and past smoking. The authors therefore consider these effects likely to be reversible (Dólleman 2013).

According to current knowledge, the number of eggs itself cannot be increased because the supply is established before birth (Wallace & Kelsey 2010). In the guidelines and studies we reviewed, we found no method that has been shown to lead to more eggs. Body mass index, waist circumference, alcohol, and exercise were not associated with age-specific AMH in the Doetinchem study (Dólleman 2013).

What else affects the AMH level

Hormonal contraception: On the combined pill, AMH was on average 17% lower (Nelson, Fertility and Sterility 2023, 42,684 women), 19% lower (Birch Petersen 2015, vaginal ring included), or 31.1% lower (Landersoe, Eur J Contracept Reprod Health Care 2020), and on the minipill 35.6% lower (same study). For the vaginal ring and the implant, the data are inconsistent, and for the hormonal IUD no clear effect was seen. The Roche group did not use contraceptives, so values on the pill only partly match the table.

Pregnancy, cycle, and smoking: In the Doetinchem study of 2,320 women from the general population, age-specific AMH in pregnant women was 17 percentiles lower than in women with regular cycles. With irregular cycles and with current pill use, it was 11 percentiles lower in each case. Current smoking was associated with AMH 4 percentiles lower, regardless of the number of cigarettes (Dólleman 2013).

PCOS: In the Roche study, 149 women with polycystic ovary syndrome had a median of 6.81 ng/ml (P5 to P95: 2.41 to 17.1 ng/ml, not broken down by age). The 2023 international PCOS guideline included AMH in adults as an alternative to ultrasound for the criterion of polycystic ovaries (Teede 2023). This is one of several criteria, and an AMH value alone does not make a diagnosis. More on this in the article Getting pregnant with PCOS.

AMH test: when it makes sense and what it costs

This section applies to Germany, Austria, and Switzerland.

Medically, the anti-Müllerian hormone test is intended mainly for the evaluation of infertility and before a planned hormone stimulation. The S2k guideline from DGGG, OEGGG, and SGGG counts AMH as part of basic hormone testing for infertility (Toth 2019, formally valid until January 2022) and recommends an evaluation after 12 months without pregnancy, or after 6 months for couples over 35. Before stimulation, the value helps with planning the dose (Nawroth 2014). As a check without a fertility evaluation, the professional societies do not recommend the test (Nawroth 2014, ASRM 2020). Which other tests belong to the evaluation and what they cost is explained in the article Fertility test for women: costs and process.

Costs in Germany

Without a medical indication, AMH is a self-pay service (IGeL) with no item number of its own in the German fee schedule for physicians. In the price lists we checked, the lab price alone was mostly around €29 to €60; Bioscientia lists €23.46 under a lower-priced item.

The example calculation includes the lab price (€29.15 to €50.28), consultation (fee schedule item 1, €10.72), and blood draw (fee schedule item 250, €5.36) at the 2.3-fold rate, plus expenses such as shipping where applicable. Before an IGeL service, the practice must inform you in text form of the expected costs (Section 630c, paragraph 3 of the German Civil Code, BGB).

For infertility, the statutory health insurance fund pays for the tests to find the cause (familienplanung.de). The practice can then bill AMH through the insurer if it justifies the medical necessity in the individual case (EBM item 32361). Some labs still list AMH as a self-pay service, so ask before the blood draw. What the insurers cover for treatment is shown in our health insurance comparison.

Costs in Austria

At Kepler University Hospital Linz, AMH costs €90 as a self-pay service, and this also applies in cases of infertility. The insurers named there pay only for endometriosis, PCOS, the onset of the menopausal transition once family building is complete, or a planned or completed operation on the ovaries (form valid from June 26, 2025). This is an example, not a general rule of the Austrian Health Insurance Fund. At synlab Austria, AMH with a blood draw costs €59.50 including VAT and is marked as a service not covered by the regional health insurance fund.

Costs in Switzerland

AMH is not named in the Analysis List (Analysenliste) of the Federal Office of Public Health (BAG), as of July 1, 2026. Coverage by mandatory basic health insurance (Grundversicherung) is therefore not assured, and insurers decide case by case. The lab analysis costs about CHF 55 (Zetlab: CHF 54.90), and with a venous blood draw at a collection site about CHF 66 including VAT.

You can find fertility clinics in Germany, Austria, and Switzerland in the clinic directory.

Frequently asked questions

Which AMH level is typical at which age?

In the Roche reference study, the median was 4.00 ng/ml at ages 20 to 24, 3.31 at 25 to 29, 2.81 at 30 to 34, 2.00 at 35 to 39, 0.882 at 40 to 44, and 0.194 ng/ml at 45 to 50. Within each age group, values vary widely, and other tests give different numbers.

Which AMH level is typical at 35, 40, and 45?

At ages 35 to 39, the median was 2.00 ng/ml (P5 to P95: 0.405 to 6.96 ng/ml); at 40 to 44, 0.882 ng/ml (0.059 to 4.44 ng/ml); and at 45 to 50, 0.194 ng/ml (P95: 1.79 ng/ml). More than half of the women aged 40 to 44 in the study were therefore below 1.0 ng/ml. These ranges describe the study group and are not a normal range for an individual woman.

Can you get pregnant with a low AMH level?

In a US study of women aged 30 to 44 with no known fertility problems, 84% of those with AMH below 0.7 ng/ml had a positive pregnancy test after 12 cycles, compared with 75% in the comparison group (Steiner 2017). The difference was not statistically significant, and a positive test was counted, not a birth. In fertility treatment in Germany in 2024, with AMH below 1.0 ng/ml, the clinical pregnancy rate per fresh transfer averaged 33.3% under age 35, 25.6% at ages 35 to 39, and 13.8% at age 40 and older (D.I.R.). These are group averages, not a personal prognosis.

Can the AMH level go up again?

A single measurement can rise: AMH varies by 20% on average over the course of a cycle (Hadlow 2016), and after stopping the pill, it rose by 53% on average in a study of 68 women (Landersoe, Reproductive BioMedicine Online 2020). The supply of eggs, by contrast, only declines after birth (Wallace & Kelsey 2010), and no method for increasing it has been proven.

What does a high AMH level mean?

A high level points to many remaining follicles (Nelson, Human Reproduction Update 2023), but it does not predict a higher chance of natural pregnancy (ASRM 2020, ACOG 2019). With AMH above 3.5 ng/ml, an average of 12.38 eggs were retrieved in fertility treatment in 2024 (D.I.R.). Before stimulation, the value helps predict a strong response and estimate the risk of overstimulation (ESHRE 2025, Nawroth 2014). In PCOS, AMH is often elevated, but the value alone does not make a diagnosis.

How much does an AMH test cost?

In Germany, the lab analysis usually costs about €29 to €60, and at a gynecologist about €45 to €66 according to an example calculation under the German fee schedule. In Austria, Kepler University Hospital Linz lists €90, and in Switzerland it is about CHF 55 to 66. With a medical indication, the health insurance fund may pay.

Are at-home AMH tests informative?

At-home tests measure AMH from finger-prick blood that is sent to a lab. In a study of 90 women that was not funded by a manufacturer, finger-prick blood averaged 3.42 ng/ml and venous blood 2.74 ng/ml; the two were not fully interchangeable (Lavadia 2026). This value does not predict pregnancy either, yet 74% of 27 sales websites examined made statements about it (Johnson 2023).

Sources

• Elecsys AMH Plus, method sheet 08818916500 V2.0, section Expected values, Roche Diagnostics, as of February 2024
• Anckaert E et al.: Multicenter analytical performance evaluation of a fully automated anti-Müllerian hormone assay and reference interval determination, Clinical Biochemistry, 2016
• Access AMH, 510(k) Decision Summary K170524, U.S. Food and Drug Administration, 2017
• Test directory entry Anti-Müllerian hormone, Ulm University Hospital, Clinical Chemistry, Rev. 004 of April 4, 2024
• Anti-Müllerian hormone, MVZ Medizin-Zentrum Dortmund, laboratory medicine, retrieved October 3, 2026
• Anti-Müllerian hormone, Endokrinologikum Hamburg, test directory, retrieved October 3, 2026
• Analyte Monograph Anti-Müllerian Hormone, Association for Clinical Biochemistry and Laboratory Medicine, January 2017
• Gassner D, Jung R: First fully automated immunoassay for anti-Müllerian hormone, Clinical Chemistry and Laboratory Medicine, 2014
• Han A et al.: Comparison of the AFIAS With Roche Elecsys and Beckman Coulter Access 2 Assays for AMH Measurement, Annals of Laboratory Medicine, 2022
• Ferguson J et al.: Establishment of a WHO Reference Reagent for anti-Müllerian hormone, Reproductive Biology and Endocrinology, 2020
• Age Group-specific Reference Intervals for the Elecsys AMH Assay in Healthy Korean Women, Annals of Laboratory Medicine, 2022
• Seifer DB, Baker VL, Leader B: Age-specific serum anti-Müllerian hormone values for 17,120 women presenting to fertility centers within the United States, Fertility and Sterility, 2011
• Nelson SM et al.: Nomogram for the decline in serum antimullerian hormone: a population study of 9,601 infertility patients, Fertility and Sterility, 2011
• Almog B et al.: Age-related normograms of serum antimullerian hormone levels in a population of infertile women: a multicenter study, Fertility and Sterility, 2011
• Kelsey TW et al.: A validated model of serum anti-mullerian hormone from conception to menopause, PLoS ONE, 2011
• Wallace WHB, Kelsey TW: Human ovarian reserve from conception to the menopause, PLoS ONE, 2010
• Anticipatory Counseling Regarding Ovarian-Factor Fertility Decline, Committee Statement No. 22, American College of Obstetricians and Gynecologists, November 2025
• Testing and interpreting measures of ovarian reserve: a committee opinion, Practice Committee of the American Society for Reproductive Medicine, December 2020
• The Use of Antimullerian Hormone in Women Not Seeking Fertility Care, Committee Opinion No. 773, American College of Obstetricians and Gynecologists, April 2019
• Steiner AZ et al.: Association Between Biomarkers of Ovarian Reserve and Infertility Among Older Women of Reproductive Age, JAMA, 2017
• Guideline on ovarian stimulation for IVF/ICSI, 2025 update, European Society of Human Reproduction and Embryology, Human Reproduction 2026
• Evidence-based guideline: unexplained infertility, European Society of Human Reproduction and Embryology, Human Reproduction 2023
• Press release on AMH testing, European Society of Human Reproduction and Embryology, July 29, 2024
• Fertility problems: assessment and treatment (NG257), National Institute for Health and Care Excellence, March 31, 2026
• Nawroth F et al.: Assessment of ovarian reserve and fertility with increasing age, joint statement of DGGEF, DGRM, and BVF, Journal of Reproductive Medicine and Endocrinology, 2014
• S2k guideline 015-085 Diagnosis and treatment before assisted reproductive treatment (Toth B et al.), DGGG, OEGGG, and SGGG, 2019
• Iliodromiti S et al.: The predictive accuracy of anti-Müllerian hormone for live birth after assisted conception, Human Reproduction Update, 2014
• Nelson SM et al.: Anti-Müllerian hormone for the diagnosis and prediction of menopause: a systematic review, Human Reproduction Update, 2023
• Hadlow N et al.: Quantifying the intraindividual variation of antimullerian hormone in the ovarian cycle, Fertility and Sterility, 2016
• Birch Petersen K et al.: Ovarian reserve assessment in users of oral contraception seeking fertility advice on their reproductive lifespan, Human Reproduction, 2015
• Landersoe SK et al.: Ovarian reserve markers in women using various hormonal contraceptives, European Journal of Contraception and Reproductive Health Care, 2020
• Landersoe SK et al.: Ovarian reserve markers after discontinuing long-term use of combined oral contraceptives, Reproductive BioMedicine Online, 2020
• Nelson SM et al.: Contraceptive-specific antimullerian hormone values in reproductive-age women: a population study of 42,684 women, Fertility and Sterility, 2023
• Dólleman M et al.: Reproductive and lifestyle determinants of anti-Müllerian hormone in a large population-based study, Journal of Clinical Endocrinology and Metabolism, 2013
• Teede HJ et al.: Recommendations from the 2023 International Evidence-based Guideline for the Assessment and Management of Polycystic Ovary Syndrome, European Journal of Endocrinology, 2023
• Yearbook 2024, German IVF Registry (D.I.R.), pp. 12 to 13, data as of April 16, 2025
• Lavadia CMM et al.: Comparison of serum AMH concentrations measured in capillary fingerstick and venous samples from Korean reproductive-aged women, Clinical and Experimental Reproductive Medicine, 2026
• Johnson A et al.: Websites Selling Direct-to-Consumer Anti-Müllerian Hormone Tests, JAMA Network Open, 2023
• German Fee Schedule for Physicians, Federal Ministry of Justice, retrieved October 3, 2026
• IGeL price list, MVZ Labor Ravensburg, May 2026
• Lab information: ovarian (residual) function, synlab MVZ Labor München Zentrum, April 2026
• Price list IGeL and private, Labor Leipzig, December 10, 2024
• Price list Mein Labor, MVZ Labor Berlin, January 2026
• Anti-Müllerian hormone, Bioscientia diagnostics app, retrieved October 3, 2026
• EBM item 32361, diego.one, retrieved October 3, 2026
• Section 630c of the German Civil Code (BGB): cooperation of the contracting parties; duties to provide information, Federal Ministry of Justice, retrieved October 3, 2026
• Fertility treatment: what is allowed and who bears the costs?, familienplanung.de (Federal Institute for Public Health), as of January 7, 2026
• AMH request form, Kepler University Hospital Linz, valid from June 26, 2025
• Private rates, synlab Austria, January 2026
• Analysis List (KLV Annex 3), Federal Office of Public Health (BAG), edition of July 1, 2026
• Legal situation in Switzerland and costs, 360° Kinderwunsch Zentrum Zürich, March 13, 2025
• Anti-Müllerian hormone (AMH), Zetlab, retrieved October 3, 2026
• Anti-Müllerian hormone (AMH) blood test, price listed by Vitalcheck (venous blood draw at a collection site, CHF 66.20 including VAT), retrieved October 3, 2026 (sales page, therefore no link)