Free Testosterone: Sex- and Cycle-Related Differences, Normal Ranges, and Optimal Levels
At a glance
- Lab category / Androgen panel
- What it measures / Unbound, biologically active testosterone (not bound to SHBG or albumin)
- Free T as share of total T / Roughly 1 to 3% in men; well under 2% in women, figures vary by assay
- Reference method / Equilibrium dialysis (gold standard) or a validated calculated free T formula
- Common pitfall / Direct "free T" immunoassays are unreliable at low concentrations
- Key clinical uses / TRT dose titration, PCOS biochemical hyperandrogenism, HRT monitoring
- Fasting required / No, but a consistent draw time (and, in women, cycle day) matters
What free testosterone is, and why total testosterone alone can mislead
Testosterone circulates in three states: bound tightly to SHBG, bound loosely to albumin, and unbound ("free"). Only the free fraction crosses cell membranes and binds the androgen receptor efficiently; the albumin-bound fraction is weakly bioavailable, and the SHBG-bound fraction is functionally inactive. Because SHBG concentration is itself a variable, it rises with aging, hyperthyroidism, liver disease, and oral estrogen, and falls with obesity, hypothyroidism, and androgen excess, two patients with identical total testosterone can have very different free testosterone, and therefore different symptom burdens. This is the core reason free testosterone is ordered as a distinct test rather than inferred from total testosterone.
How free testosterone should be measured
Equilibrium dialysis followed by mass spectrometry is widely regarded as the reference standard for free testosterone measurement, and Endocrine Society guidance on androgen testing has emphasized its accuracy problems with cheaper "direct" analog immunoassays, particularly at the low concentrations typical of women and hypogonadal men. A calculated free testosterone value, derived from total testosterone, SHBG, and albumin using a validated formula (commonly attributed to Vermeulen and colleagues), has performed well against equilibrium dialysis in published comparisons and is a reasonable substitute when dialysis is not locally available. Readers and clinicians relying on a specific numeric threshold from any single study should confirm it against the current primary literature, since exact PMID-level sourcing for several figures in earlier versions of this guidance could not be verified for this revision.
Typical ranges by sex, and why "normal" is not one number
Men
Free testosterone in adult men is commonly reported in the range of roughly 5 to 20 pg/mL by equilibrium dialysis, but the exact boundaries vary by laboratory, assay, and reference population, and they decline with age. Longitudinal cohort data (including the Massachusetts Male Aging Study) describe a gradual, roughly 1 to 2% annual decline in total testosterone beginning around age 30, with free testosterone falling somewhat faster because SHBG tends to rise with age. Any age-stratified table of exact cutoffs should be treated as illustrative rather than diagnostic; a patient's own lab reference range, tied to that lab's specific assay, takes precedence.
Free testosterone in men also varies across the day, peaking in the morning and declining by 20% or more into the afternoon in several studies of diurnal rhythm. This is why diagnostic samples in men are conventionally drawn in the morning (roughly 7 to 10 a.m.).
Women
Free testosterone in women runs several-fold lower than in men and changes across the menstrual cycle. Ovarian androgen output, and therefore free testosterone, is lowest in the early-to-mid follicular phase, rises modestly around the mid-cycle LH surge (reflecting both increased ovarian secretion and a transient dip in SHBG), and returns toward follicular-phase levels in the luteal phase. Because of this pattern, a single free testosterone value in a woman is only interpretable if the cycle day (or menopausal status and HRT regimen) is known. A mid-luteal result compared against a mid-follicular reference range, or vice versa, can produce a misleading conclusion.
Oral estrogen-containing therapy (combined oral contraceptives or oral menopausal hormone therapy) raises hepatic SHBG production substantially, which can suppress free testosterone even when total testosterone is unchanged. This is a plausible contributor to persistent low-libido symptoms in some women on oral estrogen, though a causal, quantified relationship for an individual patient cannot be asserted from population-level associations alone.
The SHBG effect: same total testosterone, different free testosterone
SHBG concentration is the main variable that determines how much of a given total testosterone pool is actually bioavailable. This matters clinically because SHBG, not testosterone, is often what shifts first with a physiologic change.
Conditions generally associated with higher SHBG (lower free T for a given total T): oral estrogen use, hyperthyroidism, cirrhosis and other chronic liver disease, and significant caloric restriction or low body weight. Aging is also associated with rising SHBG in men.
Conditions generally associated with lower SHBG (higher free T for a given total T): obesity and insulin resistance, hypothyroidism, androgen-excess states such as PCOS, and nephrotic syndrome.
In PCOS, SHBG is frequently reduced, which amplifies free testosterone even when total testosterone is only mildly elevated. This is the practical reason biochemical hyperandrogenism in PCOS evaluation is better captured by free testosterone (measured or calculated) than by total testosterone alone, a position reflected in guidance from the Androgen Excess and PCOS Society.
Free testosterone as a diagnostic anchor in PCOS
Biochemical hyperandrogenism is one of the three Rotterdam criteria for PCOS, and free testosterone is the more sensitive marker of the two testosterone measures because it reflects the low-SHBG state common in PCOS. A patient can have biochemically meaningful androgen excess with a total testosterone that stays within the normal range.
Published PCOS prevalence estimates in reproductive-age women commonly cited in the clinical literature fall in a range around 6 to 12%, though the exact figure depends on which diagnostic criteria (Rotterdam, NIH, or AE-PCOS) are applied and which population is sampled; readers should confirm current prevalence estimates against a primary epidemiologic source (for example, CDC or professional-society materials) rather than relying on a single cited number.
Specific numeric cutoffs for "elevated" free testosterone in PCOS evaluation vary by lab and assay and should be interpreted against the reporting laboratory's own reference range rather than a fixed pg/mL threshold quoted from a single paper. Direct immunoassay-based free testosterone results are generally considered less reliable for this purpose than equilibrium dialysis or a calculated value, because of assay variability at low concentrations.
Free testosterone and testosterone therapy: what titration actually depends on
Men on testosterone replacement therapy (TRT)
Endocrine Society clinical practice guidance on male hypogonadism (FDA-approved indication: testosterone therapy for men with confirmed hypogonadism due to an established medical condition) supports targeting testosterone concentrations in the normal range rather than pursuing the highest tolerated dose, and recommends biochemical monitoring in the months after starting or changing therapy, followed by periodic monitoring once stable. A specific numeric target quoted as guideline language should be treated as a paraphrase here rather than a verified direct quotation, since the exact wording could not be confirmed against the primary source for this revision.
A recognized failure mode in TRT titration: a man with low SHBG (often related to obesity or metabolic syndrome) can show a normal total testosterone while free testosterone runs high, which is associated with increased erythrocytosis risk. Relying on total testosterone alone in a patient with abnormal SHBG can therefore miss a dose that is, in effective terms, too high.
The TRAVERSE trial, a large randomized cardiovascular-safety trial of testosterone therapy in men with hypogonadism and elevated cardiovascular risk published in the New England Journal of Medicine in 2023, is the most relevant large trial evidence on cardiovascular outcomes during testosterone therapy; its detailed findings and the specific safety population it enrolled should be verified against the primary publication before being used to reassure an individual patient, since population-level trial results do not automatically transfer to every patient profile (older age, cardiovascular comorbidity, and dose all vary the applicable risk).
Women on testosterone therapy
No testosterone product currently carries FDA approval for use in women in the United States (status as of this writing; confirm current FDA labeling before relying on this for a specific patient). Use in women for symptoms such as low sexual desire is off-label. International consensus guidance on female testosterone therapy has recommended targeting free testosterone within the physiologic premenopausal female range rather than exceeding it, and has specifically cautioned against supraphysiologic dosing because long-term cardiovascular and breast safety data at higher levels are lacking. This is guideline-level caution, not a demonstrated safety threshold from a dedicated long-term safety trial.
What "optimal" does and does not mean here
"Optimal" is not the same as "highest achievable within the reference range." The available evidence supports a mid-normal target range for men on therapy and a target near premenopausal norms (not above them) for women on therapy, with higher-than-physiologic free testosterone associated with adverse effects, erythrocytosis and unfavorable lipid shifts in men, and androgenic side effects such as acne, hirsutism, and voice change in women, rather than added benefit. Observational associations between "upper-normal" free testosterone and better sexual function, lean mass, or bone density in men come from cohort studies and describe population associations, not a guarantee for an individual patient, and should not be read as license to target the top of the reference range by default.
Evidence boundary: what is established, what is plausible, what is not
Established: Free testosterone is the pharmacologically active fraction; SHBG materially determines how much testosterone is bioavailable at a given total testosterone; free testosterone in women varies meaningfully across the menstrual cycle; direct immunoassay methods for free testosterone are less accurate than equilibrium dialysis or a validated calculated value, especially at low concentrations; oral estrogen raises SHBG and can lower free testosterone.
Plausible but not firmly quantified for an individual patient: Exact pg/mL cutoffs for "optimal" free testosterone in men and women; the degree to which a specific free testosterone value predicts symptom improvement on therapy; how quickly SHBG normalizes after stopping oral estrogen-containing contraceptives.
Not established from the material available here: Precise age-stratified numeric reference intervals down to the decimal point (these vary by laboratory and assay and should not be treated as universal); long-term cardiovascular or breast-cancer safety data for testosterone therapy in women at any free testosterone level; a verified exact quotation of Endocrine Society guideline language (the paraphrase above should be checked against the current published guideline before being presented as a direct quote).
A decision framework for interpreting an unexpected free testosterone result
This framework is meant to guide the next diagnostic step, not to replace clinical judgment or a full hypogonadism or hyperandrogenism workup.
| Pattern observed | What it most often means | Reasonable next step |
|---|---|---|
| Free T low, total T normal, SHBG high | SHBG elevation (oral estrogen, hyperthyroidism, liver disease, aging) is masking a true androgen deficiency | Identify and address the SHBG driver first (for example, consider transdermal rather than oral estrogen); recheck free T before changing any androgen therapy |
| Free T low, total T low, SHBG normal | Consistent with a genuine hypogonadal state | Confirm with a second morning sample on a separate day; evaluate LH/FSH to distinguish primary from secondary hypogonadism |
| Free T high, total T normal-to-high, SHBG low | Consistent with PCOS, obesity-related SHBG suppression, or an exogenous androgen source | Screen for PCOS (with a full Rotterdam workup, not free T alone); take a careful history for supplement or exogenous androgen use |
| Free T rising unexpectedly on stable TRT dosing | Possible SHBG decline (for example, new obesity-related insulin resistance) increasing the free fraction at the same total T dose | Recheck hematocrit and lipids; consider a dose or interval adjustment rather than assuming the dose itself changed |
| Result drawn without documented cycle day (women) or draw time (men) | Result may not be comparable to the reference range applied | Redraw with cycle day or morning timing documented before acting on the number |
The organizing rule: a free testosterone number is only as interpretable as the SHBG, timing, and (in women) cycle-phase context around it. Treating an isolated free T value as self-explanatory is the most common source of misclassification described in the literature on this topic.
Practical measurement notes
Morning draws (roughly 7 to 10 a.m.) are conventional for men because of diurnal variation. For women, documenting cycle day on the requisition matters more than time of day. Direct analog immunoassays for free testosterone are widely available but are known to perform less reliably at low concentrations; when a precise result matters for a diagnostic threshold (for example, PCOS evaluation or monitoring a woman on testosterone therapy), equilibrium dialysis or a calculated free testosterone using total testosterone, SHBG, and albumin is preferable. A minimum useful panel for a new hypogonadism or hyperandrogenism workup typically includes total testosterone, free testosterone, SHBG, albumin, LH, FSH, and often prolactin and TSH, since thyroid and prolactin abnormalities can themselves alter SHBG and androgen metabolism.
A single low free testosterone value in a man is not sufficient to diagnose hypogonadism; guideline practice generally calls for at least two morning samples on separate days before starting therapy. Once therapy is started, monitoring is typically done in the first few months after initiation or a dose change, then at longer intervals once stable; women on testosterone therapy are generally monitored more frequently given how quickly small dose increases can produce supraphysiologic levels.
When to seek in-person evaluation rather than adjust based on a single number
A free testosterone result, in isolation, should not be the basis for starting, stopping, or changing hormone therapy. Symptoms consistent with androgen deficiency (persistent low libido, unexplained fatigue, loss of muscle mass) or androgen excess (new hirsutism, acne, menstrual irregularity, voice change) warrant a clinical evaluation that includes a repeat, properly timed sample and the surrounding panel described above, not a single lab value acted on alone. Rapid or severe symptoms, for example, signs suggestive of a virilizing tumor, sudden severe hirsutism, or symptoms of a testosterone-related cardiovascular event on therapy, warrant prompt medical evaluation rather than waiting for a routine follow-up lab.
Frequently asked questions
What is the difference between free testosterone and total testosterone?
Does free testosterone change during the menstrual cycle?
Why does oral estrogen lower free testosterone in women?
Is the direct free testosterone immunoassay accurate?
How does PCOS affect free testosterone?
Can symptoms occur with low free testosterone even when total testosterone is normal?
Is testosterone therapy FDA-approved for women?
A note on sourcing for this revision: The prior version of this article cited specific PubMed identifiers and journal references for numeric claims (age-stratified reference intervals, exact PCOS prevalence figures, and a direct quotation attributed to an Endocrine Society guideline). Several of these identifiers could not be verified as pointing to the correct paper for this revision, and one external link was found to point to an unrelated CDC page. Rather than retain unverified numeric citations, this draft has described the underlying evidence in general terms and flagged where a precise number or direct quotation requires confirmation against the primary literature before publication. Editorial and medical review should verify the Endocrine Society 2018 hypogonadism guideline, the Androgen Excess and PCOS Society 2009 position statement, the Global Consensus Position Statement on testosterone therapy for women (2019), and the TRAVERSE trial (NEJM, 2023) directly before any specific numeric threshold or quotation from these sources is restored to the page.