Testosterone Reference Range by Age

Normal for your age is not the same thing as healthy

There are two completely different questions hiding inside "is my testosterone normal for my age", and almost every chart on the internet answers neither of them properly. The first is descriptive: out of a hundred men your age, how many would have a lower number than you? The second is diagnostic: is this level low enough to be causing your symptoms and to warrant treatment? The first has a published answer. The second does not have an age-adjusted one, and nobody has agreed that it should.

The range your laboratory prints, and the one every guideline uses, comes from one place: Travison and colleagues in 2017 pooled four large cohorts, re-measured a subset of every cohort at the CDC by mass spectrometry to put them on a common scale, and reported the distribution in healthy nonobese men aged 19 to 39. That gave 264 to 916 ng/dL (9.2–31.8 nmol/L). It is a young-men range by construction. It is not age-adjusted, and it does not become age-adjusted by being applied to a 68-year-old.

The same paper also published something most sites either ignore or misquote: age-specific centiles, decade by decade, from 19 all the way to 99, for nonobese men and separately for all men. Those are real numbers from a primary source and the whole table is reproduced below. What they show is narrower than the headlines suggest. The median for nonobese men falls from 531 ng/dL at 19–39 to 477 at 70–79 — about 54 ng/dL across fifty years. The bottom of the distribution drops much further than the top: the 5th centile goes from 304 to 218 ng/dL, while the 95th barely moves at all, from 850 to 839.

What the paper did not do is turn those centiles into age-specific diagnostic cut-offs, and it is explicit about why that question is still open. Its authors ask directly whether the range should come from healthy young men or be age-adjusted, and answer that the issue remains unresolved. The Endocrine Society guideline that followed a year later adopted the single 264–916 ng/dL range for all adult men and said clinicians can use it for any CDC-certified assay. So this page will give you your percentile among men your age, because that is knowable, and it will not tell you that a lower number is acceptable because you are older, because nobody has established that.

One honest wrinkle, since it is in the same paper and almost nobody prints it. Travison reports two sets of numbers for the 19–39 reference sample: the observed distribution, 264–916 ng/dL between the 2.5th and 97.5th centiles, and the modelled estimates in Table 4, which give 267–929 for the same men. The range quoted everywhere, and adopted by the guideline, is the observed pair. The age-stratified table below is the modelled one, because that is the only form the older decades were published in. So the 19–39 column of that table will not match the headline range exactly, and it is not supposed to.

Two things are usually left off these charts and both move the number more than you would guess. The first is body weight: at the median, nonobese men 19–39 sit 24 ng/dL above all men 19–39 — roughly half the 54 ng/dL that fifty years of ageing costs at the same centile, and rather more than that at the bottom of the distribution. The second is the assay. These numbers only hold for a total testosterone measured on a fasting morning sample by a method calibrated to the CDC reference procedure. An uncalibrated immunoassay result is not interchangeable with them, and Travison's own conclusion was that a substantial proportion of the variation between cohorts came from assay differences rather than from the men.

Testosterone percentile by age — where a result sits among men the same age

Find your total testosterone down the left and your age across the top. The cell is the percentage of nonobese men in that age band who would have a lower result. This is a position in a distribution, not a verdict.

Every cell is computed by the same function the calculator above calls, interpolating between the nine centiles Travison 2017 published for each band. Nothing is estimated beyond the printed 2.5th and 97.5th centiles. Follow the 264 ng/dL row across — it is the harmonized diagnostic lower limit at every age. Against the modelled curve for men 19–39 it sits below 2.5th; against men 70–79 it reaches only the 7th percentile. That is the whole of what ageing fifty years buys a man holding a borderline number, and it is a lot less than "normal for your age" implies.

Total testosterone centiles by age — nonobese men (Travison 2017, Table 4)

The published age-stratified table in full, for nonobese men (N = 6,933), harmonized to the CDC reference method. ng/dL with nmol/L beneath.

Travison TG, et al. J Clin Endocrinol Metab. 2017;102(4):1161–1173, Table 4 (PMID 28324103). Model-based estimates from constrained quantile regression on the four harmonized cohorts. These are population centiles, not reference limits: the paper states the harmonized normal range as 264–916 ng/dL from the observed distribution in healthy nonobese men 19–39, and leaves the question of whether ranges should be age-adjusted explicitly unresolved. Worth noticing that the 2.5th centile sits at 218 ng/dL as late as 70–79 and only falls away at 80–99, where it reaches 157, and that the 90th and 95th centiles are flat from 40 onward.

Total testosterone centiles by age — all men, obese included (Travison 2017, Table 4)

The same table for the whole sample of 9,054 men rather than the nonobese subset. This is the more representative population and the lower one at every single cell.

All men (obese included). Travison excluded obese men from the reference sample because, in the paper's words, obesity and comorbid conditions affect circulating total testosterone and including them could distort the reference ranges. That is a defensible choice for a diagnostic range and a poor one for answering "am I typical" — most men are not in the nonobese healthy subset. Both tables are printed here so you can see how much the choice of comparison group is doing.

What a decade costs, and what body weight costs

Starting from nonobese men aged 19–39, the change in each centile when you age fifty years, when you compare against all men instead, and when you do both. Negative means lower.

All four columns are computed from the same published table. The pattern is the one Travison describes: the lower centiles fall with age and the upper ones hold roughly steady, so ageing compresses the bottom of the distribution rather than shifting the whole of it down. At the median, moving from the nonobese reference sample to all men costs 24 ng/dL without anyone getting a day older. That is not an argument that weight causes low testosterone — this is cross-sectional data and the arrow could point either way — but it is a reason to distrust any chart that stratifies by age alone.

Frequently asked questions

What is a normal testosterone level for a 50 year old man?

There is no separate diagnostic range for a 50-year-old. The harmonized reference range every major guideline uses is 264–916 ng/dL (9.2–31.8 nmol/L), derived from healthy nonobese men aged 19–39, and it is applied unchanged at every adult age. What does change is where a given number sits in the distribution: among nonobese men aged 50–59 the published centiles run 219 ng/dL at the 2.5th to 929 at the 97.5th, with a median of 477. Those are descriptive figures, not cut-offs.

Is there a real testosterone reference range chart by age?

There is one published age-stratified table worth printing, and it is Table 4 of Travison 2017 (PMID 28324103): model-based centiles for six age bands from 19–39 to 80–99, separately for nonobese men and all men, all harmonized to the CDC reference method. It is reproduced in full on this page. The decade-by-decade "normal ranges" that appear on most other sites — an average of 600–700 ng/dL in your twenties, 450–500 by fifty, an overall 300–1,000 ng/dL — are not traceable to any primary source we could find, so they are not printed here.

Does the testosterone reference range change with age?

The population distribution does; the diagnostic range does not. Among nonobese men the 5th centile falls from 304 ng/dL at 19–39 to 218 at 80–99, while the 95th centile moves only from 850 to 839. But the Endocrine Society guideline did not adopt an age-adjusted range, and Travison's authors describe the choice between a young-reference approach and an age-adjusted one as an unresolved conceptual issue. Nobody has established a lower threshold that is acceptable because a man is older — and the adjustment is smaller than it sounds: a result of 400 ng/dL is the 21st percentile among nonobese men 19–39 and only the 32nd among men 70–79, a gain of 11 percentile points for fifty years of ageing.

How much does testosterone drop per year after 30?

The commonly repeated "1% a year" comes from longitudinal cohorts, not from the harmonized reference data. The Baltimore Longitudinal Study of Aging reported an average change of −0.124 nmol/L per year (about −3.6 ng/dL per year) using radioimmunoassay. Travison's cross-sectional harmonized data show a smaller trend — the nonobese median falls from 531 to 477 ng/dL between the 19–39 and 70–79 bands, roughly 1.2 ng/dL a year averaged across the gap — and its authors note explicitly that their age trend is smaller than earlier cross-sectional reports. The two designs disagree and this page does not pick a winner.

My testosterone is normal for my age but I feel terrible. What does that mean?

It means you have been given the descriptive answer to a diagnostic question. A result of 240 ng/dL at 65 is above the 2.5th centile for nonobese men aged 60–69 (218 ng/dL) and below the harmonized lower limit of 264 ng/dL — both true at once. And "normal for your age" is doing less work than it sounds like: 240 lands at roughly the 4th percentile among men that age, not in the middle of anything. The guideline route is to confirm the number with a repeat fasting morning measurement on an accurate assay, and to diagnose hypogonadism only where symptoms and signs accompany consistently low levels. Being typical for your age group is not evidence that you are fine.

Why does my lab report a different range from this chart?

Most likely because the laboratory quotes the interval for its own assay rather than the harmonized one, and assays are not interchangeable. The CDC Hormone Standardization Program certifies total testosterone assays against a mass spectrometry reference procedure, and the Endocrine Society guideline notes that LC-MS/MS assays generally offer higher specificity, sensitivity and precision than most immunoassays, particularly at low concentrations. Travison's own conclusion was that a substantial proportion of the variation between cohorts came from assay differences. Every number on this page assumes a CDC-standardised result.

Should obese men use a different testosterone reference range?

Travison published both, which is why both are on this page. At the median, nonobese men aged 19–39 sit at 531 ng/dL against 507 for all men of the same age — a gap of 24 ng/dL, about half of what fifty years of ageing costs at the same centile (54 ng/dL). At the 2.5th centile the two are almost level: 38 ng/dL for the comparison group against 49 for ageing into your seventies. The diagnostic range comes from the nonobese sample, because including obese and comorbid men could distort a reference range. For the question "am I typical", the all-men table is the more honest comparison, and this calculator lets you switch between them.

Why is the reference range so much higher than what most US men actually measure?

Because a reference range and a population measurement are different things, and almost nobody says so. Travison's reference sample excluded obese men and men with comorbid conditions, and among nonobese men 19–39 its 10th to 90th centiles run 344–774 ng/dL. NHANES 2011–2012, a nationally representative sample of the US household population measured by standardised LC-MS/MS at the CDC, reported 10th–90th percentiles of 150–698 ng/dL in men aged 20 and over. Those two spans barely overlap at the bottom. The reference range describes healthy young men; NHANES describes everybody. Neither is wrong and neither substitutes for the other.

What does it mean to be in the 10th percentile for testosterone?

It means nine men in ten of the same age and body-weight category would have a higher result than you, and one in ten lower. For nonobese men aged 40–49 the published 10th centile is 310 ng/dL. A percentile is not a diagnosis: somebody has to be in the bottom tenth of any distribution, and being there is not in itself abnormal. It is useful as context alongside symptoms, a repeat measurement, SHBG and a free testosterone calculation — not as a result in its own right.

Sources

1. [1]Travison TG, Vesper HW, Orwoll E, et al. Harmonized Reference Ranges for Circulating Testosterone Levels in Men of Four Cohort Studies in the United States and Europe. J Clin Endocrinol Metab. 2017;102(4):1161–1173 — "Harmonized normal range in a healthy nonobese population of European and American men, 19 to 39 years, is 264 to 916 ng/dL."
2. [2]Travison TG, et al. 2017, full text — Table 4, "Model-Based Estimates of Population Centiles for Total Testosterone Concentrations (ng/dL) Based on Data From Nonobese Men (N = 6933) and in All Men (N = 9054) in the Four Harmonized Cohorts", the source of every age-stratified figure on this page
3. [3]Bhasin S, et al. Endocrine Society Clinical Practice Guideline, 2018, abstract — "We recommend measuring fasting morning total T concentrations using an accurate and reliable assay as the initial diagnostic test."
4. [4]Centers for Disease Control and Prevention, Steroid Hormones Standardization Programs — "CDC Hormone Standardization Programs assesses the analytical accuracy and reliability of testosterone and estradiol tests performed in clinical, research and public health laboratories."
5. [5]Vesper HW, Wang Y, Vidal M, Botelho JC, Caudill SP. Serum Total Testosterone Concentrations in the US Household Population from the NHANES 2011-2012 Study Population. Clin Chem. 2015;61(12):1495–1504 — "The 10th-90th percentiles of totalT values in adults (≥20 years) was 150-698 ng/dL (5.20-24.2 nmol/L) in men"
6. [6]Harman SM, Metter EJ, Tobin JD, et al. Longitudinal effects of aging on serum total and free testosterone levels in healthy men. Baltimore Longitudinal Study of Aging. J Clin Endocrinol Metab. 2001;86(2):724–731 — "an average change of -0.124 nmol/L.yr"