Free versus total testosterone, and the three ways labs get the free number
"Free testosterone" on a report can mean any of three quite different procedures. Two of them are reasonable and one is widely regarded as unreliable, and reports do not always say which you got.
Why a free number exists at all
Most circulating testosterone is bound. A large fraction is held tightly by sex hormone-binding globulin, a further fraction is held loosely by albumin, and a small percentage circulates free. The bound-to-SHBG portion is generally considered unavailable to tissue on the relevant timescale; the free portion, and to a debated degree the albumin-bound portion, are not.
That is the entire reason free and bioavailable testosterone are measured. If your SHBG is unusual in either direction, total testosterone stops tracking what is actually available — which is how two men with identical total testosterone end up with quite different free levels and quite different experiences.
The three methods, and which to trust
Equilibrium dialysis physically separates the free fraction and measures it. It is the reference method, it is more expensive and slower, and it is what the other two are validated against.2
Calculated free testosterone derives the number from total testosterone, SHBG and albumin using the law of mass action. The widely used version comes from Vermeulen and colleagues in 1999,1 and it correlates well with dialysis when the inputs are measured on reliable assays.1 It is what most panels labelled "free testosterone, calculated" are giving you.
Direct immunoassay — sometimes called analogue free testosterone — is the cheap one, and the app's registry note calls it unreliable. That is the mainstream view rather than a strong opinion.
Established clinical use That the three methods are not interchangeable is established, and the app stores which one produced a value for the same reason it does for estradiol.
The calculated number inherits its inputs
A calculated free testosterone is only as good as the total testosterone, the SHBG and the albumin that went into it. Change the SHBG assay and the calculated free value moves without anything having changed in you — which is the registry's own note on this marker, and the reason the app records the method.
Off-label or community practice There is also live scientific argument about the model itself. The binding model underlying the classic mass-action equations has been challenged, with evidence of allosteric interaction between the two SHBG binding sites, and reassessments have found discrepancies against direct equilibrium dialysis.3 This does not make calculated free testosterone useless — it remains the most practical option for most people — but it does mean a calculated value carries more uncertainty than its decimal places suggest.
Bioavailable testosterone is the free fraction plus the albumin-bound fraction, and it is usually calculated from the same three inputs, so it inherits the same error.
Where SHBG comes into it
SHBG is the lever. High SHBG binds more testosterone tightly, so a given total produces less free; low SHBG does the reverse. That is why a total testosterone result read without an SHBG alongside it is incomplete for anyone whose symptoms and numbers do not match.
The SHBG page covers what moves it. For the arithmetic itself there is a free testosterone calculator that runs the Vermeulen equation in whatever units your lab printed and shows its working.
What the app records for this marker
- Marker
- Free Testosterone
- Panel group
- hormones
- Canonical unit
- pg/mL
- Units accepted
- pg/mL (canonical)
pmol/L (× 0.288 → pg/mL)
ng/dL (× 10 → pg/mL)
ng/L (× 1 → pg/mL) - Assay methods tracked
- Direct immunoassay (
direct-immunoassay)
Equilibrium dialysis (equilibrium-dialysis)
Calculated (from SHBG) (calculated) - Generic reference range
- 8–25 pg/mL — generic reference range, male default. Your lab's printed interval takes precedence over this.
- Optimal band
- 15–25 pg/mL non-diagnostic
- LOINC
- 2991-0 — unverified seed, not checked against a vendor payload
How the app adjusts this range by sex and age
Generated by running the app's own range function against each band, so this table cannot drift from what the app flags against. The optimal column is non-diagnostic throughout.
| Who | Reference range | Optimal band |
|---|---|---|
| Male, under 40 | 9–30 pg/mL | 15–25 pg/mL |
| Male, 40–49 | 7–24 pg/mL | 12–22 pg/mL |
| Male, 50 and over | 5–21 pg/mL | 10–18 pg/mL |
| Female, all ages | 0.3–1.9 pg/mL | 0.8–1.8 pg/mL |
Compounds whose monitoring panel includes this marker
Generated by matching this marker's own aliases against the monitoring note on every compound in the app's reference, so the list is the app's, not an author's.
- Testosterone Cypionate Total T, Free T, E2 (sensitive assay), SHBG, Hematocrit, CBC, PSA, Lipids, CMP. Every 6-8 weeks adjusting; every 3-6 months stable.
- Anastrozole E2 sensitive assay, Total T, Free T every 6-8 weeks when adjusting.
- Enclomiphene Total T, Free T, LH, FSH, E2, Semen analysis (if fertility is goal), every 6-8 weeks when adjusting.
- Exemestane E2 sensitive assay, Total T, Free T, Lipid panel, Bone density with long-term use.
- Dutasteride / Finasteride Total T, Free T, DHT levels (if testing available), PSA, sexual function self-assessment monthly.
- Liothyronine (T3) Free T3, Free T4, TSH, Reverse T3, heart rate (resting), blood pressure, bone density (long-term). TSH will suppress — this is expected at therapeutic T3 doses.
- Levothyroxine (T4) Free T4, Free T3, TSH, Reverse T3 at 6-8 weeks after any dose change, then every 6-12 months stable.
- Natural Desiccated Thyroid Free T4, Free T3, TSH, heart rate. TSH typically suppresses — guide dose by Free T3 target.
- DHEA DHEA-S (primary marker), Total T, Free T, E2 (DHEA converts to both), PSA in men, lipid panel.
- Testosterone Pellets Total T, Free T, E2, Hematocrit at 4 weeks post-insertion and at trough (just before next insertion). PSA annually. Insertion site monitoring for infection or extrusion.
- Proviron Free Testosterone; SHBG; Estradiol (E2); Lipid panel; DHT
- Testosterone Propionate Total and Free Testosterone; Estradiol; Hematocrit; PSA men over 40; Lipid panel
Questions people actually ask
Which free testosterone should I ask for?
Equilibrium dialysis is the reference method if you can get it and afford it. Calculated is the practical default and is fine provided the total testosterone and SHBG behind it came from good assays and you keep using the same lab. Direct immunoassay free testosterone is the one to be sceptical of. Which is appropriate for your situation is a question for the clinician ordering it.
My total is fine but my free is low. Which do I believe?
Both, usually — that pattern is what high SHBG looks like, and it is exactly the situation free testosterone exists to reveal. Check whether an SHBG was run at the same time. If the free value was a direct immunoassay, treat it with more caution than the total.
Why does my calculated free testosterone change when I switch labs?
Because the calculation takes SHBG as an input and SHBG assays are not standardised across labs. The app records the method with the value so a change of that kind is visible as a method change rather than read as a change in you.
Sources
Every claim above that is not a definition is either labelled by evidence tier or carries a numbered reference to one of these.
- Vermeulen A, Verdonck L, Kaufman JM, A critical evaluation of simple methods for the estimation of free testosterone in serum, J Clin Endocrinol Metab 84(10):3666 (1999) The origin of the calculated free testosterone equation in general use.
- Free testosterone by direct and calculated measurement versus equilibrium dialysis in a clinical population, The Aging Male (2013) Head-to-head comparison of the three approaches in practice.
- Reassessing Free-Testosterone Calculation by Liquid Chromatography–Tandem Mass Spectrometry Direct Equilibrium Dialysis, J Clin Endocrinol Metab 103(6):2167 (2018) Challenges the binding model behind the classic equations and quantifies the discrepancy.
TherapyLog logs this marker with the unit, the reference interval your report printed and the assay method beside it, so a trend cannot silently switch methods on you.
Log your bloodworkThis calculator does the arithmetic you typed and nothing else. It does not know what is actually in your vial, whether the label is accurate, or anything about you. Confirm the vial strength and the diluent volume on your own label before you draw, and take dosing decisions to a qualified provider.