The same person can look low, normal, or borderline depending on whether the lab reports calculated free testosterone, analog free testosterone, equilibrium dialysis, or free androgen index. The difference is usually chemistry, not biology.
- Free testosterone is the small unbound fraction of testosterone, often about 1-3% of total testosterone in adults.
- Calculated free testosterone uses total testosterone, SHBG, and albumin; it is usually the practical choice when total testosterone and symptoms disagree.
- Equilibrium dialysis testosterone is the reference method many endocrinologists trust most, but it is slower, costlier, and not available in every lab.
- Analog free testosterone test results can be misleading because the assay is strongly affected by SHBG and is not recommended for many diagnostic decisions.
- Free androgen index equals 100 × total testosterone divided by SHBG, using the same molar units; it is more useful in women than in men.
- Low SHBG from obesity, insulin resistance, nephrotic syndrome, or hypothyroidism can make total testosterone look low while free testosterone is less abnormal.
- High SHBG from aging, hyperthyroidism, liver disease, some medicines, or oral estrogen can make total testosterone look acceptable while free testosterone is low.
- Morning testing matters: men should usually repeat testosterone between 7-10 am, ideally fasting, after a normal sleep night and no acute illness.
- Doctor-trusted method when results conflict is usually repeat morning total testosterone by a reliable assay plus SHBG and albumin to calculate free testosterone, with equilibrium dialysis if stakes are high.
Why free testosterone results change by method
Free testosterone results change because each method measures a different approximation of the unbound hormone. Calculated free testosterone estimates it from total testosterone, SHBG, and albumin; equilibrium dialysis physically separates it; analog free testosterone test kits often track SHBG more than true free hormone; free androgen index is a ratio, not a direct free testosterone measurement.
As of June 17, 2026, my usual clinical approach is simple: if symptoms and total testosterone disagree, I look for SHBG, albumin, timing, acute illness, medicines, and the actual free testosterone method before I label anyone hypogonadal. A total testosterone of 310 ng/dL with SHBG of 12 nmol/L is not the same clinical problem as 310 ng/dL with SHBG of 70 nmol/L.
Kantesti is an AI blood test interpretation platform that reads testosterone results alongside SHBG, albumin, LH, FSH, prolactin, thyroid markers, glucose, A1c, liver enzymes, and medication context. That pattern-based reading is the reason our reports often explain why a patient’s free versus total testosterone story looks inconsistent on the first page.
I am Thomas Klein, MD, and in clinic I have seen this exact confusion derail care: a tired 44-year-old man was told his analog free testosterone was “normal,” but his SHBG was 82 nmol/L and his calculated free testosterone was plainly low. The fix was not immediate treatment; it was a repeat 8 am panel, review of thyroid and liver markers, and a cleaner diagnosis.
What free testosterone actually represents in circulation
Free testosterone is the fraction not bound tightly to SHBG or loosely to albumin. In most adults, only about 1-3% of circulating testosterone is truly free, while roughly 30-60% is SHBG-bound and the rest is mostly albumin-bound.
SHBG is the main reason two people with the same total testosterone can feel and test very differently. Low SHBG pushes total testosterone down and often preserves calculated free testosterone; high SHBG can keep total testosterone in range while reducing free testosterone.
Age, sex, liver function, thyroid status, insulin resistance, oral estrogen exposure, anticonvulsants, and body composition all change SHBG. That is why I never interpret testosterone without checking whether the reference interval is sex-specific; our guide to sex-based lab ranges explains why a flag can be technically correct but clinically unhelpful.
Albumin matters less than SHBG in most outpatient cases, but it is not irrelevant. When albumin falls below about 3.5 g/dL in liver disease, nephrotic syndrome, malnutrition, or inflammatory illness, a calculator that assumes 4.3 g/dL can overstate the precision of calculated free testosterone.
How calculated free testosterone is produced
Calculated free testosterone is an equation-based estimate using total testosterone, SHBG, and albumin. The Vermeulen equation remains widely used because it performs reasonably well when total testosterone and SHBG are measured accurately.
Vermeulen, Verdonck, and Kaufman compared common estimation methods in 1999 and found that calculated free testosterone could approximate reference methods better than many simple ratios when good input assays were used (Vermeulen et al., 1999). The hidden weakness is obvious but often missed: a calculator cannot rescue a poor total testosterone assay or an inaccurate SHBG result.
A practical example: total testosterone 280 ng/dL with SHBG 10 nmol/L may yield a calculated free testosterone in or near range, while total testosterone 420 ng/dL with SHBG 75 nmol/L may calculate low. In our hormone reviews, that distinction often changes the next step from “replace testosterone” to “find out why SHBG is abnormal.”
Calculated free testosterone is most useful when it is interpreted with LH, FSH, prolactin, estradiol, TSH, A1c, and liver markers. Patients who want the broader pattern can compare their result against our hormone panel patterns rather than staring at one number in isolation.
Why equilibrium dialysis testosterone is often trusted
Equilibrium dialysis testosterone is often treated as the reference method because it physically separates free hormone from protein-bound hormone. It is technically demanding, slower than routine immunoassays, and still depends on careful laboratory handling.
In equilibrium dialysis, serum is placed on one side of a semipermeable membrane and buffer on the other; unbound testosterone diffuses until equilibrium is reached. The free fraction is then measured, often after extraction and mass spectrometry in higher-quality laboratories.
Doctors trust this method when the result has real consequences: borderline hypogonadism, suspected androgen excess in women, unusual SHBG values, litigation or elite sport contexts, or a patient whose symptoms do not fit routine testing. Rosner and colleagues warned in the Endocrine Society position statement that testosterone measurement is vulnerable to method-related error, especially at low concentrations (Rosner et al., 2007).
Kantesti AI flags equilibrium dialysis results differently from routine analog free testosterone because method reliability is part of clinical interpretation, not a footnote. Our clinical validation standards describe how we separate method quality from the medical meaning of the number.
Why the analog free testosterone test can mislead
The analog free testosterone test is a direct immunoassay that often does not measure true free testosterone accurately. It can be distorted by SHBG concentration, which is exactly the variable doctors are trying to account for.
This is the test I am most cautious about. A direct analog assay may look convenient because it produces a free testosterone number quickly, but convenience is not the same as validity when SHBG is low in obesity or high with aging, thyroid disease, or medication effects.
The 2018 Endocrine Society guideline advises diagnosing testosterone deficiency only in men with consistent symptoms and unequivocally low testosterone confirmed on repeat testing, and it discourages unreliable free testosterone assays when better methods are available (Bhasin et al., 2018). In plain English: a single analog free testosterone result should rarely carry the diagnosis by itself.
If your report says “free testosterone, direct” or “free testosterone analog,” ask whether your clinician can repeat total testosterone with SHBG and albumin or order equilibrium dialysis. For a broader diagnostic path, our low testosterone workup lays out the follow-up labs that usually matter more than one assay flag.
Where free androgen index helps and where it fails
Free androgen index is a ratio, not a measured free testosterone test. It is calculated as 100 × total testosterone divided by SHBG, with both values in the same molar units, usually nmol/L.
FAI can be helpful in women being assessed for androgen excess because total testosterone is low and SHBG changes can amplify symptoms. In PCOS workups, a raised FAI may align with acne, hirsutism, irregular cycles, insulin resistance, or elevated LH-to-FSH patterns.
FAI performs poorly in many men because SHBG sits in the denominator and can exaggerate the impression of androgen exposure. A man with SHBG 8 nmol/L can have a very high FAI even when true free testosterone is not high enough to explain symptoms or safety risks.
The ratio also breaks down when total testosterone is measured by a low-quality immunoassay at female-range concentrations. That is why I prefer LC-MS/MS total testosterone plus SHBG in women whenever possible, especially when reviewing PCOS hormone patterns.
Which method doctors trust when symptoms and total disagree
When symptoms and total testosterone disagree, doctors usually trust repeat morning total testosterone plus SHBG-based calculated free testosterone, and they use equilibrium dialysis when the decision is high-stakes. A lone analog free testosterone test is usually the least persuasive result.
The Endocrine Society guideline recommends confirming low testosterone with repeat morning testing and using accurate assays before diagnosing hypogonadism (Bhasin et al., 2018). In many adult men, a total testosterone below about 264 ng/dL is below the harmonized lower limit used in guideline discussions, but symptoms and repeat confirmation still matter.
Kantesti is an AI biomarker interpretation platform that treats a borderline testosterone value as a pattern question: SHBG, LH, FSH, prolactin, estradiol, TSH, A1c, ferritin, sleep clues, and medication effects all change the answer. In my experience, that prevents the common error of treating a number when the physiology says “slow down.”
I often use a three-step rule in clinic: repeat the test at 7-10 am, calculate free testosterone if SHBG is unusual, and reserve equilibrium dialysis for cases where the diagnosis will trigger long-term testosterone therapy, fertility decisions, or endocrine referral. Men comparing age-related symptoms can start with a structured andropause lab panel.
Units, ranges, and calculator inputs that change the answer
Free testosterone calculators change when units, albumin assumptions, or reference ranges are entered incorrectly. Total testosterone in ng/dL must be converted to nmol/L for many equations, and free testosterone may be reported as pg/mL, ng/dL, pmol/L, or nmol/L.
The most common conversion I see mishandled is total testosterone: ng/dL × 0.0347 equals nmol/L, and nmol/L × 28.8 equals ng/dL. For free testosterone, 1 pg/mL is approximately 3.47 pmol/L, because testosterone’s molecular weight is about 288.4 g/mol.
Albumin defaults vary by calculator. If a calculator assumes albumin 4.3 g/dL but your measured albumin is 2.9 g/dL, the result may be less reliable, especially in liver disease, kidney protein loss, or severe inflammatory states.
Reference intervals are method-specific. A free testosterone range derived from equilibrium dialysis should not be pasted onto an analog free testosterone report, just as cholesterol equations should not be interpreted like direct assays; our guide to unit changes explains why identical biology can look different on paper.
Timing and preparation can move testosterone results
Testosterone testing is timing-sensitive, especially in men under 45. Levels are usually highest in the early morning, and poor sleep, hard training, calorie restriction, alcohol, acute illness, or opioid use can lower results for days to weeks.
For most men, I prefer a repeat sample between 7-10 am, fasting if feasible, after at least one ordinary sleep night. A single 4 pm testosterone of 245 ng/dL in a sleep-deprived shift worker is not enough to diagnose chronic hypogonadism.
Exercise is tricky. Heavy resistance training can transiently raise or lower testosterone depending on timing, calorie balance, and recovery; a brutal session 12-24 hours before testing can also move CK, AST, cortisol, and inflammatory markers in ways that confuse the clinical picture.
Medication history belongs on the lab order, not hidden in the visit note. Opioids, glucocorticoids, anabolic agents, antiandrogens, some anticonvulsants, and fertility drugs can all shift testosterone or SHBG, so patients should review testosterone test prep before repeating a borderline panel.
SHBG patterns in obesity, aging, and testosterone therapy
SHBG explains many testosterone contradictions in obesity, aging, and testosterone therapy monitoring. Low SHBG is common with insulin resistance and obesity, while SHBG often rises with aging, thyroid excess, liver disease, and some medications.
In obesity, total testosterone often falls before true gonadal failure is present. I have seen men lose 8-12% body weight and raise total testosterone by 100-200 ng/dL, largely through improved insulin resistance and higher SHBG rather than a sudden change in the testes.
In older men, the opposite pattern is common: SHBG rises, total testosterone may look deceptively reassuring, and calculated free testosterone falls. That is why symptoms such as low libido, reduced morning erections, anemia, low trauma fracture, or loss of muscle should be matched against free testosterone, LH, and safety labs.
Kantesti AI also treats testosterone therapy results as timing-dependent. An injection trough of 320 ng/dL and a peak of 1,100 ng/dL can belong to the same person, so our analysis often points users toward obesity and testosterone or therapy-specific timing before they overreact.
Free testosterone testing in women needs extra caution
Free testosterone testing in women is harder because concentrations are much lower than in men. Standard immunoassays often struggle at female-range testosterone levels, so LC-MS/MS total testosterone plus SHBG-based calculation is usually more informative.
Many healthy premenopausal women have total testosterone roughly in the 15-70 ng/dL range, but lab intervals vary widely by age, cycle status, contraception, and assay. Free testosterone may be reported in low pg/mL values, where small analytical errors become clinically large.
Oral estrogen-containing contraception can raise SHBG substantially, sometimes above 150 nmol/L, which lowers calculated free testosterone even when total testosterone looks unchanged. Conversely, insulin resistance can lower SHBG and raise FAI, which is one reason androgen symptoms and metabolic markers belong in the same conversation.
For women with acne, hirsutism, cycle changes, scalp hair thinning, or suspected PCOS, I prefer total testosterone by LC-MS/MS, SHBG, DHEA-S, 17-hydroxyprogesterone when indicated, TSH, prolactin, A1c, and lipids. Our guide to women’s testosterone ranges goes deeper into age and cycle timing.
How Kantesti interprets free testosterone in context
Kantesti interprets free testosterone by checking the method, units, SHBG, albumin, sex-specific ranges, timing, and related endocrine markers together. A single free testosterone flag is treated as a clue, not a diagnosis.
Kantesti is an AI-powered blood test analysis tool used by more than 2M people across 127+ countries, and our neural network is designed to spot patterns that patients often miss in PDF reports. For testosterone, the common hidden pattern is low SHBG plus insulin resistance, not isolated androgen failure.
Our AI checks whether total testosterone was likely measured by immunoassay or LC-MS/MS, whether free testosterone was calculated, analog, or dialysis-based, and whether albumin was measured or assumed. The AI technology guide explains how method metadata and biomarker relationships are handled without turning the report into a black box.
There are limits. Kantesti AI can flag inconsistency, recommend sensible follow-up questions, and explain why two methods disagree, but it cannot examine you, assess fertility goals, or replace clinician judgment; that is also why our article on AI interpretation limits is worth reading before making treatment decisions.
What to ask your doctor before acting on the result
Before acting on a free testosterone result, ask which method was used, whether the test was repeated in the morning, and whether SHBG and albumin were included. Those three questions prevent many unnecessary prescriptions and missed diagnoses.
I tell patients to bring the full report, not just a screenshot. A “normal” free testosterone result without SHBG, albumin, collection time, and assay method is often too thin to guide treatment, especially if symptoms are significant.
Thomas Klein, MD would usually ask about libido, erections or sexual pain, fertility plans, energy, strength, sleep apnea, hot flushes, cycle changes, acne, hair pattern changes, medicines, and prior anabolic exposure before drawing conclusions. If testosterone is repeatedly low, LH and FSH help separate primary gonadal failure from pituitary or functional suppression.
Clinician oversight matters, particularly before testosterone therapy, fertility treatment, or stopping medication. Kantesti’s physician review process is supported by our medical advisory board, and patients unsure about conflicting labs can also read when to seek a second opinion.
Research publications and transparency behind our approach
Transparent methods matter because free testosterone interpretation depends on assay quality, unit handling, and clinical context. Kantesti publishes technical and health-data reports so patients and clinicians can see how our blood test interpretation work is evaluated.
Kantesti Ltd. (2026). AI Blood Test Analyzer: 2.5M Tests Analyzed | Global Health Report 2026. Zenodo. DOI. ResearchGate: publication search. Academia.edu: archive search. The internal global report provides broader deployment context across countries, languages, and common lab panels.
Kantesti Ltd. (2026). RDW Blood Test: Complete Guide to RDW-CV, MCV & MCHC. Zenodo. DOI. ResearchGate: publication search. Academia.edu: archive search. Although RDW is not a testosterone marker, the publication shows the same interpretive principle: method, units, trend, and clinical context beat isolated flags.
Thomas Klein, MD reviews this topic through a cautious endocrine lens: a calculator result is useful only when the inputs are trustworthy and the clinical question is clear. Readers comparing multiple hormone markers can use our biomarker guide as a map, but treatment decisions still belong with a qualified clinician who knows the patient.
Frequently Asked Questions
Why is my free testosterone different on two lab reports?
Free testosterone can differ across lab reports because calculated free testosterone, equilibrium dialysis, analog immunoassays, and free androgen index do not measure the same thing. A change in SHBG from 20 nmol/L to 70 nmol/L can shift the interpretation even when total testosterone changes little. Unit differences also matter: total testosterone in ng/dL must be converted to nmol/L for many calculators. Ask the lab or clinician which method was used before comparing numbers.
Is calculated free testosterone accurate enough?
Calculated free testosterone is often accurate enough for routine clinical interpretation when total testosterone, SHBG, and albumin are measured with reliable assays. The Vermeulen method is widely used and generally performs better than simple ratios when SHBG is abnormal. It is less dependable if total testosterone was measured poorly, albumin was assumed during severe illness, or units were entered incorrectly. In high-stakes cases, equilibrium dialysis may be preferred.
Is equilibrium dialysis testosterone the best test?
Equilibrium dialysis testosterone is commonly regarded as the reference method because it physically separates unbound testosterone from protein-bound testosterone. It is not perfect, because temperature, sample handling, dialysis conditions, and downstream measurement still affect accuracy. It is usually most useful when SHBG is very high or low, when female-range testosterone needs careful evaluation, or when treatment decisions are major. Many routine cases can start with repeat morning total testosterone plus SHBG-based calculation.
Should I trust an analog free testosterone test?
An analog free testosterone test should be interpreted cautiously because it may be strongly influenced by SHBG rather than true free testosterone. This matters when SHBG is low, such as in obesity or insulin resistance, or high, such as with aging, thyroid excess, liver disease, or oral estrogen use. A single analog result should not diagnose testosterone deficiency or androgen excess by itself. A better follow-up is usually total testosterone, SHBG, albumin, and calculated free testosterone, with equilibrium dialysis if needed.
What is free androgen index and is it the same as free testosterone?
Free androgen index is not the same as measured free testosterone. It is calculated as 100 × total testosterone divided by SHBG, using the same molar units, usually nmol/L. FAI can help screen androgen excess in women, especially in PCOS assessment, but it is less reliable in men and in people with very low SHBG. It should be read as a ratio that reflects binding protein effects, not a direct hormone concentration.
What should I do if total testosterone is normal but free testosterone is low?
If total testosterone is normal but free testosterone is low, the first step is to check SHBG, albumin, timing, and the free testosterone method. High SHBG, often above about 60 nmol/L in men, can make total testosterone look acceptable while calculated free testosterone is reduced. Repeat testing between 7-10 am is usually sensible, especially if the first sample was later in the day or during illness. Your clinician may also check LH, FSH, prolactin, thyroid markers, liver enzymes, A1c, and medication effects.
Can I use a free testosterone calculator at home?
You can use a free testosterone calculator at home for education, but it should not be used alone to diagnose or treat a hormone disorder. The calculator needs accurate total testosterone, SHBG, and albumin, and the units must be entered correctly. A wrong conversion, such as mixing ng/dL and nmol/L, can move a result from low to normal without any biological change. Use the result as a structured question for your clinician, not as a prescription decision.
Get AI-Powered Blood Test Analysis Today
Join over 2 million users worldwide who trust Kantesti for instant, accurate lab test analysis. Upload your blood test results and receive comprehensive interpretation of 15,000+ biomarkers in seconds.
📚 Referenced Research Publications
Klein, T., Mitchell, S., & Weber, H. (2026). AI Blood Test Analyzer: 2.5M Tests Analyzed | Global Health Report 2026. Kantesti AI Medical Research.
Klein, T., Mitchell, S., & Weber, H. (2026). RDW Blood Test: Complete Guide to RDW-CV, MCV & MCHC. Kantesti AI Medical Research.
📖 External Medical References
⚕️ Medical Disclaimer
This article is for educational purposes only and does not constitute medical advice. Always consult a qualified healthcare provider for diagnosis and treatment decisions.
E-E-A-T Trust Signals
Experience
Physician-led clinical review of lab interpretation workflows.
Expertise
Laboratory medicine focus on how biomarkers behave in clinical context.
Authoritativeness
Written by Dr. Thomas Klein with review by Dr. Sarah Mitchell and Prof. Dr. Hans Weber.
Trustworthiness
Evidence-based interpretation with clear follow-up pathways to reduce alarm.