HOMA2-IR, HOMA2-%B & HOMA2-%S Explained

Direct answer: HOMA2 is a nonlinear computer model, not a stand-alone blood test. From compatible fasting glucose with insulin or C-peptide, it estimates insulin sensitivity (HOMA2-%S), the reciprocal insulin-resistance index (HOMA2-IR), and fasting beta-cell activity (HOMA2-%B). These outputs may add context in selected healthcare-professional or research settings, but they do not diagnose insulin resistance, prediabetes, diabetes, reactive hypoglycemia, pancreatic disease, or a diabetes type.^{1, 2, 6}

For accepted screening, diagnostic, classification, and monitoring pathways, use the parent guide: Diabetes and Prediabetes Blood Tests: A1C, Glucose, Insulin, and C-Peptide. This page has a narrower purpose: explaining HOMA2’s three labels, inputs, interpretation, and limitations.

Educational note: Review any proposed HOMA2 calculation—and the laboratory results entered into it—with a qualified healthcare professional. Oxford states that its calculator is intended for healthcare professionals and does not replace formal medical assessment.²

Key Takeaways

• HOMA2 is calculated from paired fasting laboratory measurements; it is not measured directly in blood.
• HOMA2-%S and HOMA2-IR express the same modeled sensitivity dimension in opposite directions. They are not two independent findings.
• HOMA2-%B estimates fasting beta-cell activity within the model. It does not measure beta-cell mass, health, survival, or future reserve.
• HOMA2 and the older HOMA1/HOMA-IR approximation are related but not interchangeable.
• There is no universal HOMA2 cutoff that separates normal from abnormal or diagnoses insulin resistance.
• Collection timing, fasting conditions, specimen type, assay, units, medicines, illness, and software compatibility all matter.
• Current Oxford guidance describes HOMA2 for people not receiving insulin therapy. C-peptide does not automatically make HOMA2 valid during exogenous insulin treatment.³
• There is no automatic HOMA2 retesting schedule. Repeat only when a healthcare professional has defined the question, timing, and decision the result could inform.

Use This Guide Within the Ulta Lab Tests Knowledge Center

If laboratory testing is new to you, start with The Complete Guide to Lab Tests and Blood Work. Then use How to Read and Understand Your Lab Results for values, units, flags, reference intervals, decision thresholds, and trends. Direct-Access Lab Testing: How It Works and What to Expect explains preparation, collection, results, and responsible follow-up.

On This Page

• What HOMA2 estimates
• Three labels, two model dimensions
• HOMA2 versus HOMA-IR
• Required laboratory inputs
• How to interpret results
• What HOMA2 cannot diagnose
• When HOMA2 is the wrong tool
• Testing, preparation, and compatibility
• Frequently asked questions
• Medical references

What Is HOMA2?

HOMA stands for Homeostasis Model Assessment. It represents the feedback relationship between glucose and insulin during a fasting, basal, steady-state period. The model asks what combination of insulin sensitivity and beta-cell activity could plausibly produce the paired fasting concentrations entered into the software.

HOMA2 is an updated, nonlinear computer model. Oxford describes it as an estimate of steady-state beta-cell function and insulin sensitivity, expressed relative to a reference population. Those estimates may correspond with—but are not necessarily equivalent to—results from dynamic methods such as clamp studies or glucose-tolerance testing.¹

Oxford uses the label “beta-cell function” for HOMA2-%B. In patient-facing language, modeled fasting beta-cell activity is less likely to be overread. HOMA2-%B describes activity required by the model under the conditions measured; it is not a picture of the pancreas and does not count living beta cells.

HOMA2 output	What it estimates	How direction is usually described	What it does not establish
HOMA2-%S	Insulin sensitivity relative to the model’s reference calibration	Higher values represent greater modeled sensitivity; lower values represent less modeled sensitivity	A clamp measurement, treatment target, or universal normal range
HOMA2-IR	Insulin resistance as the reciprocal view of HOMA2-%S	Higher values correspond to lower HOMA2-%S; lower values correspond to higher HOMA2-%S	An independent third dimension or a stand-alone diagnosis
HOMA2-%B	Fasting beta-cell activity relative to glucose and the model calibration	Higher or lower values must be read beside glucose, insulin or C-peptide, medicines, and modeled sensitivity	Beta-cell mass, pancreatic health, future reserve, disease stage, or diabetes type

The reference calibration helps standardize the model. It does not make 100% a universal personal target or a pass/fail boundary.

Three Labels, Two Model Dimensions

HOMA2-%S and HOMA2-IR move in opposite directions because they describe one modeled sensitivity result. Within the model, HOMA2-IR is the reciprocal of HOMA2-%S, expressed as HOMA2-IR = 100 ÷ HOMA2-%S. Counting both as separate evidence would double-count the same dimension. HOMA2-%B is the second modeled dimension.

HOMA2 Versus the Older HOMA-IR Formula

“HOMA-IR” and “HOMA2-IR” are often used as though they mean the same thing. They do not.

The older HOMA1 approximation can be calculated with a simple equation. A commonly used version is fasting insulin multiplied by fasting glucose and divided by 405 when glucose is reported in mg/dL. That equation is not a HOMA2 formula. HOMA2 uses nonlinear software; Oxford does not publish a valid hand-calculation equation for it.^{2, 4}

Feature	Older HOMA1 approximation	HOMA2 model
Calculation	Linear approximation formula	Iterative nonlinear computer model
Common outputs	HOMA-IR and an older HOMA-%B approximation	HOMA2-%S, HOMA2-IR, and HOMA2-%B
Inputs	Commonly fasting glucose and insulin	Fasting glucose with compatible insulin, specific insulin, or C-peptide input
Interpretation	Study cutoffs vary by population and assay	Thresholds also vary; input method, specimen, assay, software version, and comparison population must be documented
Interchangeability	A HOMA1 result cannot be relabeled as HOMA2	A HOMA2 result should identify its input type and software context

The HOMA-IR Calculation (Insulin Resistance) Panel provides fasting glucose and insulin for the older calculation described on its product page. It is not a HOMA2 panel and should not be presented as producing HOMA2-IR, HOMA2-%B, or HOMA2-%S.

Which Laboratory Results Are Used for HOMA2?

HOMA2 uses paired fasting glucose plus a compatible insulin, specific-insulin, or C-peptide result. The measurements should come from the same fasting, steady-state period. A Glucose Test and an Insulin Test are the most commonly discussed pair. A C-Peptide Test addresses endogenous insulin secretion and may be used for a different modeling question.

Possible input pair	Why it may be considered	Important limitations
Fasting glucose + fasting insulin	Commonly used for modeled sensitivity and beta-cell activity	Insulin secretion is pulsatile; insulin assays differ; exogenous insulin and insulin-stimulating medicines may make interpretation unsuitable
Fasting glucose + fasting C-peptide	May provide a secretion-focused estimate using endogenous C-peptide	Kidney function affects C-peptide clearance; C-peptide-based and insulin-based estimates are not automatically interchangeable; current HOMA2 guidance still excludes people on insulin therapy
Both paired sets, professionally directed	May answer distinct sensitivity and secretion questions	More testing is not automatically better; each added test needs a defined purpose and compatible collection, specimen, assay, units, and software

The Specimen and Assay Warning Matters

Oxford states that the insulin version of the calculator was designed for glucose and insulin measured simultaneously in plasma. Correctly collected and handled serum or plasma glucose may be usable, but serum insulin should be documented as a limitation unless the laboratory has established an assay-specific serum-to-plasma conversion. Product availability alone does not establish HOMA2 compatibility.²

Before treating any results as model inputs, the interpreting professional should confirm:

• paired collection during the same fasting period;
• whether the patient was in a reasonably stable fasting state;
• serum or plasma specimen type;
• assay method and units;
• whether insulin, specific insulin, or C-peptide was entered;
• current calculator version, licensing, accepted input type, and software implementation;
• medicines, injected insulin, acute illness, pregnancy, kidney function, and other factors that could alter the measurements or their meaning.

Current Oxford version guidance describes HOMA2 use in people with or without diabetes who are not on insulin therapy.³ A direct C-peptide measurement can be useful in selected diabetes-classification questions, but that separate clinical use does not validate a HOMA2 calculation during insulin therapy. Never stop or change insulin or another medicine for testing without explicit medical direction.

How to Interpret HOMA2 Results Without Overcalling Them

There is no absolute HOMA2 value that separates normal from abnormal for every person. Oxford notes that HOMA thresholds depend on the glucose, insulin, and C-peptide assays used.² A cutoff from one research paper may not apply to a different laboratory, population, specimen, medication context, or software implementation.

Start with the underlying results, not the index alone. Confirm the correct patient, collection time, fasting conditions, units, flags, method, and relevant medicines. Then ask whether the model result agrees with accepted glycemic tests, symptoms, history, and trends. The How to Read and Understand Your Lab Results guide provides a structured review process.

Observed model result	A reasonable question to ask	Do not conclude from this alone
Higher HOMA2-IR with lower HOMA2-%S	Were fasting conditions, input compatibility, assay method, medicines, illness, and the comparison framework appropriate?	That insulin resistance has been definitively diagnosed, graded, or staged
Higher HOMA2-%B	Could fasting secretion be compensating for lower modeled sensitivity, or could medicine, assay, or collection factors be contributing?	That the pancreas is healthier, contains more beta cells, or has unlimited reserve
Lower HOMA2-%B	Does the estimate agree with simultaneous glucose, C-peptide, medicines, symptoms, and accepted diabetes tests?	That beta cells have failed or a specific diabetes type is present
Change over time	Were the fasting conditions, specimen approach, assay, software, and medication context comparable?	That the change proves treatment success or disease progression
Conflict with A1C or glucose	Could preanalytic variation, red-cell factors, acute illness, medicines, kidney function, or an unsuitable input explain the mismatch?	That HOMA2 should override accepted diagnostic testing

What HOMA2 Can—and Cannot—Tell You

HOMA2 may help	HOMA2 cannot establish by itself
Estimate fasting steady-state insulin sensitivity and beta-cell activity from compatible paired inputs	A diagnosis of insulin resistance, prediabetes, or diabetes using a universal cutoff
Support research comparisons or selected professional physiologic assessment	Post-meal glucose handling, reactive hypoglycemia, or response to a glucose challenge
Add context when fasting glucose and insulin or C-peptide are interpreted together	Type 1, type 2, type 3c, monogenic, or gestational diabetes classification
Support cautious trend analysis when methods and conditions are comparable	Beta-cell mass, pancreatic damage, beta-cell health, or future secretory reserve
Provide a model output for a defined research or locally validated framework	The same information as a clamp study, oral glucose tolerance test, continuous glucose monitor, imaging study, or clinical examination

HOMA2 Is Not a Diabetes Diagnostic Test

The American Diabetes Association’s 2026 Standards of Care identify A1C and plasma-glucose pathways for diagnosing diabetes. HOMA2 is not among those diagnostic criteria.⁶ An A1c Test estimates longer-term glycemic exposure; fasting glucose measures one point in time; insulin and C-peptide answer narrower physiologic questions.

The 2023 diabetes laboratory guideline does not recommend routine insulin or proinsulin testing for most people with diabetes or diabetes risk. It supports targeted C-peptide use in selected ambiguous classification situations.⁷ For screening thresholds, confirmation, classification, and monitoring, return to Diabetes and Prediabetes Blood Tests: A1C, Glucose, Insulin, and C-Peptide. For the broader cardiometabolic pattern, use Insulin Resistance as an Early-Warning System: Lab Patterns That Reveal Metabolic Strain.

When HOMA2 May Add Context

HOMA2 is most defensible when a healthcare professional or research protocol defines:

• the specific question the model is meant to address;
• why accepted glucose-based testing does not answer that question by itself;
• the compatible input pair, specimen, assay, units, and software;
• the comparison population or locally validated framework;
• the limitations created by medicines and clinical conditions; and
• what decision, if any, could change because of the result.

Potential settings include research comparisons, selected healthcare-professional assessment of fasting physiology, and carefully standardized follow-up. NIDDK notes that healthcare professionals may not test for insulin resistance and that insulin-resistance testing is used primarily in research.⁸ HOMA2 should therefore not be marketed as a routine requirement for everyone with fatigue, weight concerns, a family history of diabetes, or an otherwise normal fasting glucose result.

When HOMA2 Is the Wrong Tool

Do not use a fasting steady-state model to answer a dynamic, classification, or urgent-care question.

• Post-meal symptoms or suspected reactive hypoglycemia: HOMA2 does not measure a meal response or diagnose reactive hypoglycemia.
• Acute illness or unstable glucose: Infection, major stress, severe hyperglycemia, hypoglycemia, or another non-steady-state condition can make the model inappropriate.
• Current insulin therapy: Oxford’s current version history describes the model for people not receiving insulin therapy. C-peptide does not remove that restriction.³
• Insulin secretagogues or other influential medicines: A medicine can change fasting insulin or HOMA2-%B without showing improved beta-cell health.
• Diabetes classification: HOMA2 cannot determine diabetes type. Classification may require history, examination, accepted glucose testing, direct C-peptide, islet autoantibodies, pancreatic evaluation, imaging, or genetics.
• A pass/fail wellness score: No universal cutoff supports labeling each result normal, optimal, early, late, or severe.
• A substitute for symptoms and clinical care: A model output cannot rule out an important condition or explain symptoms by itself.

Urgent-care note: Do not wait for routine HOMA2 testing if you have confusion, fainting, seizures, deep or rapid breathing, repeated vomiting, severe dehydration, chest pain, marked weakness, or very high or very low glucose readings. Seek prompt professional or emergency evaluation.

Testing Options, Preparation, and Compatibility

Individual Tests Versus Paired Panels

Option	What it provides	Potential use	Main limitation
Separate Glucose Test and Insulin Test	Two individual results	May provide a paired set if collected together under compatible conditions	Separate orders do not guarantee HOMA2 compatibility
Insulin and Glucose Panel	Fasting glucose and fasting insulin in one panel	Convenient paired measurements for a defined glucose–insulin question	The panel does not diagnose insulin resistance and should not be assumed to produce HOMA2 outputs
HOMA-IR Calculation (Insulin Resistance) Panel	Glucose, insulin, and the older HOMA-IR calculation	May support the older method when that is the intended calculation	It is not HOMA2 and must not be relabeled as HOMA2-IR, HOMA2-%B, or HOMA2-%S
Add a C-Peptide Test or A1c Test	A targeted secretion measure or longer-term glycemic context	Useful when each result answers a separate question	More biomarkers do not create a single composite diagnosis or establish HOMA2 compatibility

How to Prepare for Paired Fasting Testing

Follow the current instructions for every item at the time of testing; preparation requirements can change.

1. Define the question first. Decide whether the goal is accepted diabetes screening, the older HOMA-IR calculation, a professionally calculated HOMA2 estimate, or a separate C-peptide question.
2. Use the intended paired inputs. Collect glucose with insulin or C-peptide during the same fasting period. Do not combine measurements from different days as though they represent one steady state.
3. Follow the longest applicable preparation requirement. Use only water during the stated fasting interval unless the current product instructions or a healthcare professional directs otherwise.
4. Document relevant conditions. Note recent illness, unusual exercise, poor sleep, alcohol use, major dietary change, pregnancy, and medication or supplement changes.
5. Do not stop medicines on your own. Ask the prescribing professional whether the medicine makes HOMA2 unsuitable or changes preparation.
6. Confirm compatibility before calculation. Verify specimen type, assay, units, accepted input type, software version, licensing, and whether the intended method is HOMA1 or HOMA2.
7. Standardize any justified repeat. Use comparable collection timing, fasting conditions, specimen, laboratory method, software, and medication context where feasible.

The Direct-Access Lab Testing guide explains how to select a focused test, prepare, complete collection, receive results, and plan responsible follow-up.

Questions to Ask a Healthcare Professional

• What question would HOMA2 answer that accepted glucose or A1C testing does not?
• Are these paired fasting results compatible with the current HOMA2 software?
• Were specimen type, units, assay, fasting duration, and collection timing appropriate?
• Was the model run with insulin or C-peptide, and is that input appropriate for the intended output?
• What reference population or locally validated framework is being used?
• Could medicine exposure, insulin therapy, acute illness, pregnancy, kidney function, or another factor distort the estimate?
• Does the HOMA2 result agree with glucose, A1C, symptoms, medicines, and previous results?
• Would an accepted diagnostic or dynamic test answer the question more directly?
• Would repeating the calculation under comparable conditions change a decision?
• Which symptoms or findings require routine follow-up, specialist evaluation, or urgent care?

Frequently Asked Questions

What is HOMA2-IR?

HOMA2-IR is the model’s reciprocal expression of insulin sensitivity. It is derived from compatible paired fasting measurements entered into HOMA2 software. It is not a directly measured analyte and has no universal diagnostic cutoff.

What does HOMA2-%S mean?

HOMA2-%S estimates insulin sensitivity relative to the model’s reference calibration. Higher HOMA2-%S corresponds to lower HOMA2-IR, and lower HOMA2-%S corresponds to higher HOMA2-IR. The two outputs are not independent confirmation.

What does HOMA2-%B mean?

HOMA2-%B estimates fasting beta-cell activity in relation to glucose and the model calibration. It does not report the percentage of beta cells remaining, beta-cell mass, pancreatic health, or future insulin-producing reserve.

What is a normal HOMA2-IR, HOMA2-%B, or HOMA2-%S?

There is no universal normal-versus-abnormal threshold. Assay method, specimen, fasting conditions, medicines, software, and the comparison population all matter. A study-specific cutoff should not be applied to an individual unless the interpreting professional has established that it fits the same method and context.

Can HOMA2 diagnose insulin resistance?

No universal HOMA2 cutoff diagnoses insulin resistance in routine care. HOMA2 may provide model-based context for a defined professional or research question, but it does not replace clinical assessment.

Can HOMA2 diagnose prediabetes or diabetes?

No. Prediabetes and diabetes use accepted A1C and plasma-glucose pathways, with confirmation when required. HOMA2 is not part of the ADA diagnostic criteria and should not override an accepted diagnostic result.⁶

Can HOMA2 diagnose reactive hypoglycemia?

No. Reactive hypoglycemia is a post-meal, dynamic question. HOMA2 describes a fasting steady state and does not show what happens after food or a glucose challenge.

Can HOMA2 distinguish type 2 diabetes from type 3c diabetes?

No. HOMA2 cannot classify diabetes type or diagnose pancreatic disease. Classification may require specialist-directed history, examination, direct laboratory testing, imaging, or genetics.

Can HOMA2 be used while taking insulin?

Current Oxford version guidance describes HOMA2 for people not receiving insulin therapy.³ Using C-peptide instead of insulin does not automatically validate HOMA2 in this setting. Do not alter insulin for testing.

Can I calculate HOMA2 myself?

HOMA2 cannot be calculated with the older hand formula. Oxford licenses the current software for healthcare-professional use and says it is not intended for patients except in consultation with a trained medical adviser.² A website labeled “HOMA2 calculator” should not be assumed to implement the validated model.

How often should HOMA2 be repeated?

There is no automatic interval. Repeat only when a defined question remains, enough time has passed for a meaningful change, collection and methods can be made comparable, and the result could alter a decision. Diabetes monitoring follows its own evidence-based schedule.

Explore Related Diabetes, Insulin Resistance, and Metabolic Health Guides

• Diabetes and Prediabetes Blood Tests: A1C, Glucose, Insulin, and C-Peptide covers screening, confirmation, classification, and monitoring.
• Insulin Resistance as an Early-Warning System: Lab Patterns That Reveal Metabolic Strain covers the broader laboratory pattern involving glucose, insulin, lipids, liver, kidney, and inflammation context.
• Insulin Resistance and Type 2 Diabetes: Risk, Testing, and Prevention explains risk, progression, screening, and prevention.
• Insulin Resistance Diet: Foods, Meals, and Lab Tests focuses on food selection and meal planning rather than HOMA2 mechanics.
• Metabolic Health Blood Tests: Weight Loss, Insulin Resistance, and GLP-1 Monitoring covers weight, cardiometabolic risk, and individualized monitoring during GLP-1 care.

Laboratory Inputs and Related Tests Discussed in This Guide

HOMA2-IR, HOMA2-%B, and HOMA2-%S are calculated outputs—not individual Ulta laboratory products.

• Possible model inputs requiring compatibility confirmation: Glucose Test, Insulin Test, and C-Peptide Test
• Related paired testing: Insulin and Glucose Panel
• Different calculation—not HOMA2: HOMA-IR Calculation (Insulin Resistance) Panel
• Accepted glycemia context: A1c Test

How Ulta Lab Tests May Help

When a defined clinical question supports fasting glucose, insulin, C-peptide, or A1C testing, eligible adults may review available individual tests and panels online, follow current preparation instructions, complete collection, and discuss the results with a healthcare professional. Availability does not mean that every test is appropriate or that any result is automatically compatible with HOMA2.

Conclusion

HOMA2-IR, HOMA2-%B, and HOMA2-%S can describe aspects of fasting glucose–insulin physiology, but their value depends on a valid question, compatible paired inputs, the correct software, and restrained interpretation. HOMA2-IR and HOMA2-%S are reciprocal views of one sensitivity dimension. HOMA2-%B estimates modeled fasting activity—not beta-cell health or reserve.

No universal HOMA2 cutoff diagnoses insulin resistance, prediabetes, diabetes, reactive hypoglycemia, pancreatic disease, or diabetes type. Use accepted glucose and A1C pathways for screening and diagnosis, reserve HOMA2 for an appropriate professional or research context, and interpret it beside the underlying results, symptoms, history, medicines, collection details, assay, related tests, and trends.

Medical References

1. University of Oxford, Radcliffe Department of Medicine. HOMA Calculator. Accessed September 1, 2026.
2. University of Oxford, Radcliffe Department of Medicine. HOMA Calculator Frequently Asked Questions. Accessed September 1, 2026.
3. University of Oxford, Radcliffe Department of Medicine. HOMA Calculator Version History. Accessed September 1, 2026.
4. Wallace TM, Levy JC, Matthews DR. Use and Abuse of HOMA Modeling. Diabetes Care. 2004;27(6):1487–1495.
5. Levy JC, Matthews DR, Hermans MP. Correct Homeostasis Model Assessment (HOMA) Evaluation Uses the Computer Program. Diabetes Care. 1998;21(12):2191–2192.
6. American Diabetes Association Professional Practice Committee for Diabetes. Diagnosis and Classification of Diabetes: Standards of Care in Diabetes—2026. Diabetes Care. 2026;49(Suppl 1):S27–S49.
7. Sacks DB, Arnold M, Bakris GL, et al. Guidelines and Recommendations for Laboratory Analysis in the Diagnosis and Management of Diabetes Mellitus. Diabetes Care. 2023;46(10):e151–e199.
8. National Institute of Diabetes and Digestive and Kidney Diseases. Insulin Resistance & Prediabetes. Last reviewed March 2025. Accessed September 1, 2026.
9. Jones AG, Hattersley AT. The Clinical Utility of C-Peptide Measurement in the Care of Patients with Diabetes. Diabetic Medicine. 2013;30(7):803–817.

Update and corrections note: Before publication, assign named medical and laboratory-methodology reviewers, document their approval, add verified editorial and corrections-policy links, directly confirm each product landing page and canonical URL, and set the visible update date and structured-data dateModified to the actual deployment date.

Written by: John R.
Medical review: To be assigned and documented before publication
Laboratory-methodology review: To be assigned and documented before publication
Originally published: July 16, 2026
Revision prepared: September 1, 2026
Next scheduled review: Within 12 months after documented medical review, or sooner if guidance or laboratory methods change
Disclosure: Ulta Lab Tests offers laboratory-testing services and links to tests and panels discussed in this guide. This content is educational and does not provide individual diagnosis or treatment.