How to Test Testosterone Properly

A test of testosterone means a morning venous blood draw, ideally between 7 and 10 a.m., measuring total testosterone and repeated on a second separate day. The following are measured alongside it.

• sex hormone-binding globulin (SHBG)
• albumin
• luteinizing hormone (LH)
• follicle-stimulating hormone (FSH)
• prolactin

A single total testosterone value is close to uninterpretable on its own. Within-person day-to-day variation is large, and testosterone follows a diurnal rhythm that peaks in the early morning. The fraction of that total which is biologically available depends entirely on SHBG, which varies several-fold between people. If you order one number and get it in the afternoon after a bad night of sleep, you have measured noise. The panel below is what we would order, and the rest of this piece explains why each piece is there and how the assay chemistry changes what the number means.

The panel we would order

Total testosterone is the anchor measurement, and it should be measured by liquid chromatography-tandem mass spectrometry (LC-MS/MS) rather than by a direct immunoassay wherever you can choose. Immunoassays are calibrated for mid-range adult male concentrations and lose accuracy at the low end, which is precisely the range where the result changes what you do next. Mass spectrometry is the reference method and handles low concentrations and structurally similar steroids far better than antibody-based platforms, which is the core argument reviewed in the assay literature.¹ Quest and Labcorp both offer LC-MS/MS total testosterone as a distinct orderable, usually labeled “testosterone, total, LC/MS-MS” or similar, and it costs a few dollars more than the immunoassay version.

SHBG and albumin come next, because they let you calculate free testosterone. Roughly 98% of circulating testosterone is bound, mostly to SHBG with high affinity and to albumin with low affinity, so the unbound fraction is what determines the signal reaching tissue. The Vermeulen equation takes total testosterone, SHBG, and albumin and returns calculated free and bioavailable testosterone. Use that calculation rather than a direct free testosterone immunoassay or an analog free testosterone test. Direct free testosterone assays are the least reliable part of a standard hormone panel and are known to disagree substantially with equilibrium dialysis, the reference method.¹ If you have access to equilibrium dialysis free testosterone, order it instead of calculation.

LH and FSH tell you where in the axis a low value originates. Low total testosterone with elevated LH points toward the testis. Low testosterone with low or low-normal LH points toward the pituitary or hypothalamus. Prolactin is there because a markedly elevated prolactin suppresses the axis and has a specific cause worth finding. Estradiol, measured by LC-MS/MS (the immunoassay is unreliable at male concentrations), is useful if you want a fuller picture of aromatization. A complete blood count with hematocrit and a comprehensive metabolic panel give you the context that testosterone acts in, and hematocrit in particular is the biomarker most sensitive to exogenous androgen exposure over time.

Timing, preparation, and the sources of noise you control

Testosterone peaks in the early morning in most men and declines through the day, with the amplitude of the rhythm larger in younger men. Draw before 10 a.m. If you cannot, note the time and hold it constant across draws so that your own serial values are comparable even if they are not comparable to a reference interval built on morning samples.

Several conditions lower measured testosterone, sometimes markedly and transiently.

• acute illness
• sleep restriction
• heavy recent exercise
• opioid use
• high-dose glucocorticoids
• substantial caloric restriction

Fasting matters modestly: glucose ingestion suppresses testosterone for a couple of hours, so fast for the draw. Do not test during or within a few weeks of an acute illness. Guidelines are consistent on requiring two separate morning measurements before treating a value as real, and in our experience the second draw reclassifies a meaningful fraction of borderline results.

Assay traceability matters if you compare numbers across labs. Reference intervals differ between platforms, and a value of 310 ng/dL on one immunoassay is not the same measurement as 310 ng/dL by LC-MS/MS on a CDC-traceable calibration. If you are building a longitudinal series, stay with one lab and one method, and record the method name in your own records.

What the number does and does not tell you

Serum testosterone reports on circulating concentration, which is not the same as intratesticular concentration. The two can diverge, notably under exogenous androgen or certain fertility treatments. Serum 17-hydroxyprogesterone has been studied as an indirect readout of intratesticular testosterone, and it correlates with intratesticular testosterone in ways serum testosterone does not.² The distinction matters for anyone thinking about fertility, and it is a conversation for a urologist rather than a self-interpreted panel.

Low testosterone travels with a broader biomarker signature. In a cohort of men undergoing evaluation, low plasma testosterone was associated with elevated high-sensitivity cardiovascular biomarkers, including markers of inflammation and myocardial stress.³ Separately, low serum testosterone in older men has been associated with higher all-cause mortality over follow-up in a population-based cohort.⁴ Associations of this kind constrain how you should read a low value: it is a signal worth investigating with a clinician, and it does not by itself establish causation or imply that raising the number changes the outcome.

In the other direction, you may want to know whether a testosterone concentration in a given person is exogenous or endogenous. The serum concentration alone is a weak discriminator, and longitudinal biomarker panels are what carry the information. Work on detecting transdermal testosterone administration in women with normal cycles found that multiple biomarkers tracked over time outperformed single static measurements.⁵ The same logic applies to your own baseline: a personal series of ten morning values under controlled conditions is far more informative than a single value compared against a population interval.

Reading a result, concretely

Start with total testosterone by LC-MS/MS from two morning fasting draws. If both are comfortably mid-range and you have no symptoms, the hormone axis is not where your question lies, and further testosterone testing adds little. If both are low, compute free testosterone from total, SHBG, and albumin, and look at LH and FSH to localize the problem. If total is low but SHBG is also low (common with obesity, insulin resistance, hypothyroidism, or nephrotic states), calculated free testosterone may be normal and the low total is partly a binding-protein finding. The reverse pattern, high SHBG with borderline total, occurs with aging, hyperthyroidism, and some anticonvulsants, and can leave free testosterone lower than the total suggests.

Symptoms are part of the interpretation and cannot be replaced by a number. The features below are the ones most often reported, but each has many non-hormonal causes and none is specific.

• reduced morning erections
• lowered libido
• loss of body hair
• reduced spontaneous energy
• declining muscle mass with stable training
• depressed mood

Take the panel plus the symptom history to a clinician, preferably an endocrinologist or urologist. Let them decide what follows, if anything. A low testosterone result should never lead you to self-administer androgens: exogenous testosterone suppresses the axis and impairs spermatogenesis, and the decision requires a physician. Pharmacologically induced changes in testosterone also shift common blood biomarkers on short timescales. Any subsequent panel has to be read in that light.⁶

Questions people also ask

Is testosterone included in a full blood test? No. A standard complete blood count and comprehensive metabolic panel contain no sex hormones. Testosterone, SHBG, LH, and FSH all have to be ordered explicitly.

Is TSH the same as testosterone? No. TSH is thyroid-stimulating hormone, a pituitary hormone that regulates the thyroid. Thyroid status does influence SHBG and therefore the free fraction of testosterone, so a TSH is reasonable to have alongside a hormone panel, but the two measure entirely different systems.

Why do some doctors hesitate to test testosterone? Because a single non-fasting, non-morning total testosterone generates a large number of borderline results. Those results lead to repeat testing and sometimes to treatment that was never indicated. The hesitancy is about test discipline rather than about the analyte. If you present with specific symptoms and ask for a properly timed panel, most clinicians will order it.

At what age should a man get testosterone checked? There is no age-based screening recommendation, and population screening of asymptomatic men is not advised. Testing is driven by symptoms or by specific conditions. That said, a baseline measured in your twenties or thirties is genuinely useful later, because your own prior values are a better comparator than a population reference interval.

Can I get a testosterone test at a pharmacy? Retail and direct-to-consumer options exist, and saliva or fingerstick collection kits are widely sold. We would avoid them for a first evaluation: most use immunoassay methods, collection timing is often uncontrolled, and salivary testosterone correlates imperfectly with serum. Saliva does carry other useful signals in this domain, including oxidative stress markers studied in relation to hypogonadism.⁷ For the primary measurement, order a morning venous LC-MS/MS draw.

Footnotes

1. Amin S. Herati, Cenk Cengiz, Dolores J. Lamb. Assays of Serum Testosterone. Urologic Clinics of North America, 2016. https://doi.org/10.1016/j.ucl.2016.01.003 ↩ ↩²

2. Thiago Fernandes Negris Lima, Premal Patel, Ruben Blachman-Braun, et al. Serum 17-Hydroxyprogesterone is a Potential Biomarker for Evaluating Intratesticular Testosterone. Journal of Urology, 2020. https://doi.org/10.1097/ju.0000000000001016 ↩

3. Alexander W. Pastuszak, Taylor P. Kohn, Joel Estis, et al. Low Plasma Testosterone is Associated with Elevated Cardiovascular Disease Biomarkers. The Journal of Sexual Medicine, 2017. https://doi.org/10.1016/j.jsxm.2017.06.015 ↩

4. Gail A. Laughlin, Elizabeth Barrett-Connor, Jaclyn Bergstrom. Low Serum Testosterone and Mortality in Older Men. The Journal of Clinical Endocrinology & Metabolism, 2008. https://doi.org/10.1210/jc.2007-1792 ↩

5. Olivier Salamin, Raul Nicoli, Tobias Langer, et al. Longitudinal evaluation of multiple biomarkers for the detection of testosterone gel administration in women with normal menstrual cycle. Drug Testing and Analysis, 2021. https://doi.org/10.1002/dta.3040 ↩

6. K. Barbara Sahlin, Indira Pla, Aniel Sanchez, et al. Short-term effect of pharmacologically induced alterations in testosterone levels on common blood biomarkers in a controlled healthy human model. Scandinavian Journal of Clinical and Laboratory Investigation, 2019. https://doi.org/10.1080/00365513.2019.1689429 ↩

7. Mitsuko Yasuda, Hisamitsu Ide, Kumiko Furuya, et al. Salivary 8-OHdG: A Useful Biomarker for Predicting Severe ED and Hypogonadism. The Journal of Sexual Medicine, 2008. https://doi.org/10.1111/j.1743-6109.2008.00821.x ↩