TRAb (TSH Receptor Antibody) Unit Converter

IU/L and U/L are the same number for TRAb, and mIU/L is a thousand times either. Relabelling the number is trivial; the hard part is that TRAb is a polyclonal, functionally mixed antibody population, so two assays reporting the same unit are not reporting the same quantity.

Last updated 18 Sep 2026

TRAb converter

IU relabelling only

TSH receptor antibody (TRAb)

IU/L and U/L are numerically identical. mIU/L is the IU/L figure × 1,000. There is no molar unit and no conversion to a TSI bioassay percentage — see the assay table below.

Assay cut-offTwo third-generation automated assays in routine use, both reporting IU/L, with manufacturer cut-offs of 1.75 IU/L (an electrochemiluminescence competition assay using labelled M22 monoclonal antibody) and 3.3 IU/L (a fluoroenzyme immunoassay). Choose the one your laboratory uses, or take the cut-off from the report — a TRAb read against the wrong assay’s cut-off can change the answer.

4.50IU/LExample

TRAb 4.5 IU/L

Three labels, one number, and two things that are not conversions

1 IU/L = 1 U/L (identical)
1 IU/L = 1,000 mIU/L, so mIU/L ÷ 1,000 = IU/L
IU/L → pmol/L, or → a TSI bioassay percentage: no conversion exists

IU/L = U/L: the same unit with and without the "international". A laboratory that drops the I has not changed the calibration or the number, and both appear on reports for this analyte
mIU/L: a thousand milli-international-units make an international unit, so mIU/L is the IU/L figure times a thousand. A TRAb of 4.5 IU/L is 4,500 mIU/L. It is unusual for TRAb but common enough on thyroid reports generally that a misread here is plausible
no molar unit: TRAb is a population of immunoglobulins, not a compound. There is no grams-per-mole for a polyclonal antibody response, so no pmol/L or nmol/L is offered and any site that offers one has invented it
not a TSI result: a thyroid-stimulating immunoglobulin bioassay reports cAMP generation as a percentage of basal activity, which is a biological effect and not another unit for a binding titre. A TSI bridge immunoassay does report IU/L, but it detects only stimulating antibodies, so its IU/L is a different quantity again
what the IU is worth: less than it looks. The international standard for thyroid stimulating antibody was value-assigned on receptor binding assays alone, with the explicit note that it does not provide continuity of unitage for bioassays, which read roughly twofold higher. The unit does not travel between assay classes

Worked example

TRAb 4.5 IU/L
4.5 IU/L is 4.50 U/L — the same unit written two ways
4.5 × 1,000 = 4,500 mIU/L
Above both assay cut-offs in the selector: 2.6 times the 1.75 IU/L cut-off of one and 1.4 times the 3.3 IU/L cut-off of the other, from one number
That ratio is what matters in pregnancy, where fetal surveillance is advised at about three times the upper cut-off — so the same 4.5 IU/L is near that trigger on one assay and well short of it on the other
It cannot be expressed in pmol/L, and it cannot be turned into a TSI bioassay percentage

Two third-generation assays, one unit, cut-offs almost twofold apart

Both report IU/L and both are third-generation automated assays, and the head-to-head study’s conclusion is that "given the different calibration standards used, Elecsys and EliA results are not directly comparable". Notice where the disagreement lives: negative results agree almost perfectly, positive ones agree in fewer than four cases in five. So a TRAb used as a yes-or-no test for Graves’ disease is reasonably robust, and a TRAb titre followed over time or compared against a multiple-of-cut-off threshold is only meaningful within one assay on one analyser. If a patient changes laboratory mid-course, the series restarts.
	Electrochemiluminescence competition assay	Fluoroenzyme immunoassay
Manufacturer’s cut-off	1.75 IU/L	3.3 IU/L
Measuring range	0.8–40 IU/L	—
Principle	Competition: patient antibodies inhibit binding of labelled M22 monoclonal antibody to the TSH receptor	Fluoroenzyme immunoassay on a bench autoanalyser
Calibrated against	NIBSC 1st International Standard 90/672	A different standard — which is the stated reason the two are not comparable
97.5th centile, healthy subjects	1.22 IU/L	—
97.5th centile, thyroid disease without Graves’	1.53 IU/L	—
Agreement with the other assay	Spearman correlation 0.725; agreement 99.5% for negative results and 78.5% for positive ones	As stated: "the titers from both tests were shown not to be comparable as they showed systematic and proportional differences"

TRAb, TSI bridge, TSI bioassay: three tests, three quantities

This is the strongest statement of the problem available, and it comes from the standard itself rather than from a comparison study: the institute that defines the International Unit for this analyte had to assign two different potencies depending on the assay class, and said so. A binding assay and a bioassay are measuring different physical things about the same serum — how much immunoglobulin occupies the receptor, and what happens when it does — and the same titre can represent stimulating, blocking or functionally neutral antibodies in any mixture. A reference laboratory series found that TSI assays "correlated better with clinical history and other biomarkers of Graves disease than the TRAb assay", which is what you would expect if the clinically relevant fraction is the stimulating one and TRAb counts all of it.
Test	What it detects	What it reports
Third-generation TRAb (also called TBII, TSH binding inhibitory immunoglobulin)	All antibodies that bind the TSH receptor and block the labelled monoclonal — "indiscriminately measure TSHR autoantibodies (stimulating or nonstimulating)"	IU/L
TSI bridge immunoassay	Only stimulating immunoglobulins, using "2 TSHR chimeras harbouring binding sites specific to stimulating immunoglobulins", independent of biological activity	IU/L — the same unit, a different measurand
TSI bioassay	Net stimulating activity, by cAMP production in cultured cells expressing the human TSH receptor	A percentage of basal activity
What the international standard says about this	The 2nd International Standard for thyroid stimulating antibody was "value-assigned on the basis of receptor binding assays alone", at 0.113 IU per ampoule, to preserve continuity with the previous standard "for these assays alone"	Bioassays measured roughly twofold higher potency, so establishing it "would not represent a formal continuity of unitage for the calibration of bioassays"

The two things a TRAb is actually for

The pregnancy indication is the one where the units and the assay matter most, because the rule is a ratio to a cut-off rather than a number. Three times 1.75 IU/L is 5.25; three times 3.3 IU/L is 9.9. A TRAb of 6 IU/L therefore triggers fetal surveillance on one assay and not on the other, from the same serum. That is not a reason to distrust the rule — it is the reason the rule was written as a multiple — but it does mean the cut-off has to come off the report rather than out of a textbook. Read a TRAb alongside the TSH converter and the free T4 converter, because the antibody explains the mechanism and the hormones describe the state.
Use	What guidance says
Establishing that thyrotoxicosis is Graves’ disease	The 2016 ATA hyperthyroidism guideline lists measurement of TRAb among the tests indicated when "the diagnosis is not apparent based on the clinical presentation and initial biochemical evaluation", noting that "TRAb is cost effective because if it is positive it confirms the diagnosis of the most common cause of thyrotoxicosis". A negative result does not exclude it.
Deciding whether antithyroid drugs can be stopped	The same guideline suggests measuring TRAb before stopping therapy "because it aids in predicting which patients can be weaned from the medication, with normal levels indicating greater chance for remission", with elevated end-of-therapy levels predicting relapse and low or undetectable levels associated with a substantially higher chance of permanent remission.
Predicting fetal and neonatal thyroid dysfunction	TRAb is IgG and crosses the placenta, so it can act on the fetal thyroid regardless of whether the mother is currently thyrotoxic — including in a mother cured by surgery or radioiodine. The ATA advises testing "at the time pregnancy is confirmed" in women with current or past Graves’ disease, repeating "at weeks 18-22" if elevated, and a third test "in weeks 30-34 if the second test was elevated".
The pregnancy threshold	Fetal surveillance is advised "for women whose TRAb levels are three times the upper cutoff limit". A reference-laboratory review renders the same rule as increased monitoring "when TRAb levels exceed 2 to 3 times the ULN". Both figures are given here because they differ, and because both are expressed as multiples of a cut-off rather than as an absolute IU/L — which is the only way such a threshold can work when two assays’ cut-offs differ by nearly twofold.

TRAb against the other two thyroid antibodies

All three share the property that results from different assays cannot be compared, and all three are reported in arbitrary or international units against a manufacturer’s calibration rather than in a molar unit. TRAb is nonetheless the odd one out: anti-TPO and anti-Tg are markers of an autoimmune process, while TRAb is the mechanism of the disease. That is why it has a pregnancy indication the other two do not, and why a titre in the third trimester changes what happens to a newborn.
Antibody	What a positive result tells you	Its page
TSH receptor antibody (TRAb)	The receptor is being occupied by autoantibody — in a thyrotoxic patient, that Graves’ disease is the cause. It is the only one of the three that is directly pathogenic and the only one that acts on the fetus.	this page
Thyroid peroxidase antibody (anti-TPO)	Autoimmune thyroid disease is present. Common in the general population, not specific to any one diagnosis, and not a measure of thyroid function.	anti-TPO antibody converter
Thyroglobulin antibody (anti-Tg)	Mostly that a thyroglobulin result may be unreliable. Its main role is as the companion test that decides whether a thyroglobulin measurement can be believed.	thyroglobulin antibody converter

A polyclonal antibody, an international unit that is not one thing, and one result that matters for two people

The conversion is over immediately. International units per litre and units per litre are the same unit with and without a word, and milli-international-units per litre is that figure multiplied by a thousand. There is no measured constant anywhere, no molar unit, and no arithmetic that will turn a TRAb into anything else. If relabelling the number is what you came for, it is done above, and everything below is about why the number is harder to use than it looks.

Start with what is being measured. TRAb is not a molecule; it is a polyclonal antibody response against the thyrotropin receptor, and the antibodies in it are functionally heterogeneous. Some stimulate the receptor and cause Graves’ hyperthyroidism, some block it and can cause hypothyroidism, and some bind without doing either. A third-generation TRAb assay counts them together: it measures how much patient immunoglobulin displaces a labelled monoclonal antibody from the receptor, which means it indiscriminately detects stimulating and non-stimulating autoantibodies alike. So a single titre does not tell you what the antibodies are doing, only that the receptor is occupied. Two patients with the same number can have opposite thyroid function, and one patient can shift from a stimulating to a blocking predominance over time while the titre barely moves.

Which is why the unit is less portable than the letters suggest. Two third-generation automated assays in everyday use, both reporting IU/L, have manufacturer cut-offs of 1.75 and 3.3 — nearly twofold apart. A head-to-head comparison found a Spearman correlation of 0.725 and concluded that the titres are not comparable because they show systematic and proportional differences, attributing that directly to the different calibration standards used. The pattern of disagreement is informative: negative results agreed in 99.5% of cases and positive results in 78.5%, so TRAb is a fairly robust yes-or-no test and a poor quantitative one across platforms. Older binding-inhibition assays and cell-based bioassays widen the gap further, because a bioassay reads out cAMP generation — a biological effect expressed as a percentage of basal activity — rather than binding, and a bridge immunoassay reports IU/L while detecting only the stimulating subset. Three tests, two of them in the same unit, three different quantities.

The definitive statement of this comes from the standard rather than from any study. When the second international standard for thyroid stimulating antibody was established, its potency was assigned on the basis of receptor binding assays alone, at 0.113 international units per ampoule, specifically to preserve continuity with the previous standard for those assays; bioassays estimated roughly twofold higher potency, and the documentation states that the new standard would not represent formal continuity of unitage for the calibration of bioassays. An international unit exists to make results comparable. For this analyte, the body that defines it has recorded that it does not do so across assay classes. Nothing a unit converter can do will fix that, and the honest thing for one to do is say it.

Despite all that, TRAb is a genuinely useful test for two questions. The first is causal: in a thyrotoxic patient whose clinical picture is not diagnostic, a positive TRAb confirms Graves’ disease and spares a scan, which is why the ATA lists it among the first-line discriminating tests and calls it cost-effective for that purpose. A negative result does not exclude Graves’ disease. The same guideline also suggests measuring TRAb before stopping antithyroid drugs, because a normal level predicts a substantially better chance of remission and an elevated one predicts relapse — one of the few places where the titre itself changes a decision.

The second question is about somebody who has not been born yet, and it is the reason a TRAb can matter in a woman whose own thyroid function is entirely normal. TRAb is immunoglobulin G and crosses the placenta, so it can act on the fetal thyroid in a mother whose Graves’ disease was cured years earlier by surgery or radioiodine, and in a mother rendered euthyroid by drugs. The American Thyroid Association’s schedule is to test at the time pregnancy is confirmed in any woman with current or past Graves’ disease, repeat at 18 to 22 weeks if the first result is elevated, and test a third time at 30 to 34 weeks if the second is elevated — timed so that a rising titre is identified before delivery. Fetal surveillance is advised when the level reaches about three times the upper cut-off, and a reference-laboratory review renders the trigger as two to three times the upper limit of normal. Note that the threshold is a multiple of the cut-off and not an absolute concentration, which is the only sensible way to write it given the assay spread: three times 1.75 is 5.25 IU/L and three times 3.3 is 9.9 IU/L, so a result of 6 crosses the trigger on one platform and not on the other. Take the cut-off from the report. A neonate exposed to high maternal TRAb can be thyrotoxic at birth or in the days afterwards, usually transiently as the maternal antibody clears, and that is the outcome the whole testing schedule exists to anticipate — see also the neonatal thyroid screening interpreter.

Frequently asked questions

Is IU/L the same as U/L for TRAb?

Yes, exactly. They are the same unit written with and without the word "international", and a laboratory that uses one label rather than the other has not changed the calibration or the number. mIU/L is different only by a factor of a thousand: a TRAb of 4.5 IU/L is 4,500 mIU/L.

Can I compare a TRAb result from one laboratory with one from another?

Not quantitatively. Two third-generation automated assays reporting IU/L have manufacturer cut-offs of 1.75 and 3.3 IU/L, and a head-to-head study found their titres are not comparable because they show systematic and proportional differences arising from different calibration standards. Negative results agreed in 99.5% of cases and positive results in only 78.5%, so a positive-or-negative answer travels between laboratories reasonably well and a titre does not. If a patient changes laboratory, treat the series as starting again.

What is a normal TRAb?

It depends on the assay and must come from the report. One electrochemiluminescence assay uses a cut-off of 1.75 IU/L, with 97.5th centiles of 1.22 IU/L in healthy subjects and 1.53 IU/L in people with thyroid disease other than Graves’; a fluoroenzyme immunoassay uses 3.3 IU/L. A negative TRAb does not exclude Graves’ disease, and a positive one is not a measure of how severe it is.

Is TRAb the same as TSI?

No. A third-generation TRAb assay measures all antibodies that bind the TSH receptor, stimulating or not, by how much they displace a labelled monoclonal antibody. A TSI bioassay measures only net stimulating activity, as cAMP production in cells expressing the receptor, and reports a percentage of basal activity. A TSI bridge immunoassay detects only stimulating immunoglobulins and reports IU/L — the same unit as TRAb, a different quantity. There is no conversion between any of them.

Why can the same TRAb titre mean different things in different patients?

Because the antibody population is functionally mixed. Some anti-receptor antibodies stimulate, some block and some do neither, and a binding assay counts them together. So the titre reports how much of the receptor is occupied, not what the occupancy is doing — which is why two patients with identical numbers can have opposite thyroid function, and why TSI assays have been found to correlate better with the clinical picture of Graves’ disease than TRAb does.

What TRAb level matters in pregnancy?

The rule is a multiple of the assay’s cut-off, not an absolute figure. The American Thyroid Association advises fetal surveillance when TRAb reaches about three times the upper cut-off limit, and a reference-laboratory review renders the trigger as two to three times the upper limit of normal. Testing is advised when pregnancy is confirmed in any woman with current or past Graves’ disease, repeated at 18 to 22 weeks if elevated, and again at 30 to 34 weeks if the second result is elevated. Because TRAb is IgG and crosses the placenta, this applies even to a mother whose own thyroid function is normal after surgery or radioiodine.

Why is there no molar unit for TRAb?

Because TRAb is not a compound. It is a polyclonal population of immunoglobulins of varying specificity and function, so there is no grams-per-mole to divide a mass concentration by and no mass concentration to divide. The result is an activity against an international reference preparation, which is as close to an absolute quantity as this measurand gets — and, as the reference preparation’s own documentation records, not close enough to travel between assay classes.

Related calculators

References

Roche Diagnostics. Elecsys Anti-TSHR method sheet. Reports IU/L; measuring range "0.8-40 IU/L"; cut-off 1.75 IU/L; 97.5th centiles 1.22 IU/L in healthy subjects and 1.53 IU/L in thyroid disease without Graves’; competition principle with labelled M22 monoclonal antibody; "standardized against the NIBSC … 1st IS 90/672 Standard".
Anti-TSH receptor antibodies (TRAb): comparison of two third generation automated immunoassays broadly used in clinical laboratories and results interpretation. PLoS One. 2022. doi:10.1371/journal.pone.0270890. Cut-offs 1.75 and 3.3 IU/L; Spearman 0.725; agreement 99.5% negative and 78.5% positive; "given the different calibration standards used, Elecsys and EliA results are not directly comparable".
National Institute for Biological Standards and Control. 2nd International Standard for Thyroid Stimulating Antibody, 08/204 — instructions for use. "Value-assigned on the basis of receptor binding assays alone with an assigned unitage of 0.113 IU per ampoule which will reflect formal continuity of unitage with 90/672 for these assays alone"; establishing it "would not represent a formal continuity of unitage for the calibration of bioassays".
TSH receptor antibody test utilization patterns from a national reference laboratory: TRAb, TSI, or both? J Clin Endocrinol Metab. 2026;111(1):e92. Third-generation TRAb assays "indiscriminately measure TSHR autoantibodies (stimulating or nonstimulating)"; TSI bioassays "specifically measure stimulating autoantibodies"; "TSI assays correlated better with clinical history and other biomarkers of Graves disease than the TRAb assay"; monitoring "when TRAb levels exceed 2 to 3 times the ULN".
Ross DS, Burch HB, Cooper DS, et al. 2016 American Thyroid Association guidelines for diagnosis and management of hyperthyroidism and other causes of thyrotoxicosis. Thyroid. 2016;26(10):1343–1421. doi:10.1089/thy.2016.0229. Recommendation 1 on TRAb where "the diagnosis is not apparent based on the clinical presentation and initial biochemical evaluation"; recommendation 21 on measuring TRAb before stopping antithyroid drug therapy.
Alexander EK, Pearce EN, Brent GA, et al. 2017 guidelines of the American Thyroid Association for the diagnosis and management of thyroid disease during pregnancy and the postpartum. Thyroid. 2017;27(3):315–389. doi:10.1089/thy.2016.0457. The testing schedule and the three-times-the-upper-cutoff trigger are quoted from the Association’s own rendering of this guideline in Clinical Thyroidology for the Public, vol 10 issue 5, because the guideline’s thyrotoxicosis section could not be retrieved directly.

Medical Disclaimer: The tools and content provided here are for educational and reference purposes only. They are not intended to substitute for professional medical advice, diagnosis, or treatment. Clinical decisions should always be based on the comprehensive assessment of a qualified healthcare professional.