Health information to help you understand symptoms. It isn't a diagnosis.
Check symptoms

Research deep dives

How type 2 diabetes is diagnosed

Last reviewed November 17, 2026. OwieClear editorial team. Waiting for clinician review. Not medical advice. Talk with your doctor about your own care.

Type 2 diabetes is diagnosed on blood tests, with numbers that are fixed and internationally agreed. There are four acceptable tests and any one of them can make the diagnosis.

The detail that matters most, and that is most often skipped, is that an abnormal result in someone without symptoms must be confirmed before the label is applied.

Related pages: Type 2 diabetes

The bottom line

What we know
  • A1C of 6.5 percent or higher is diagnostic.
  • Fasting plasma glucose of 126 mg/dL or higher is diagnostic.
  • A 2-hour value of 200 mg/dL or higher on an oral glucose tolerance test is diagnostic.
  • A random glucose of 200 mg/dL or higher with classic symptoms is diagnostic.
  • Without symptoms, an abnormal result needs confirmation, either by repeating the same test or by a second different test.
What is still open
  • How often A1C and fasting glucose disagree, and which to believe when they do.
  • The right A1C threshold in populations where it performs differently.
  • Whether continuous glucose monitoring should have a role in diagnosis.

The story, step by step

  1. 1. The four tests ()
  2. 2. Confirmation ()
  3. 3. Which type ()
  4. 4. When A1C misleads ()
  5. 5. What else is checked ()
What the evidence labels mean
GuidelineAn official recommendation from an expert group after weighing all the evidence.
Review of many studiesCombines the results of many studies (a systematic review or meta-analysis).
Randomized trialPeople were randomly given a treatment or a comparison, the best way to test cause and effect.
Large observational studyFollows many people over time. Shows strong links, but not always cause and effect.
Small or early studyA small study, case series, or first results. Needs to be confirmed.
Lab or animal studyDone in cells or animals. May not apply to people.
Part 1 of 5

Thresholds, exactly

Review of many studies

What happens

Any one of four results makes the diagnosis.

What the research shows

  • A1C: 6.5 percent or higher is diabetes. 5.7 to 6.4 percent is prediabetes. Below 5.7 is normal.
  • Fasting plasma glucose: 126 mg/dL or higher is diabetes. 100 to 125 is prediabetes. Fasting means no food for at least eight hours.
  • Oral glucose tolerance test, 2-hour value after a 75 gram glucose drink: 200 mg/dL or higher is diabetes. 140 to 199 is prediabetes.
  • Random plasma glucose of 200 mg/dL or higher, together with classic symptoms of thirst, frequent urination and unexplained weight loss, is diabetes without further testing.
  • The tests do not always agree with one another, and the A1C and glucose tests identify overlapping but not identical groups.
  • A1C needs no fasting and is less affected by a stressful day, which is why it is often used first, but it is the least direct measure of the four.

Why it matters

Knowing the actual numbers lets a person read their own result rather than waiting to be told what it means.

What's still uncertain

  • Which test to prefer when two disagree.
  • Whether the fasting threshold of 126 is set at the right place for all populations.

Read more

Next: Confirmation

Part 2 of 5

One abnormal test is not a diagnosis

Review of many studies

What happens

Repeat, unless symptoms are unmistakable.

What the research shows

  • In a person without symptoms, a single abnormal result should be confirmed before diabetes is diagnosed.
  • Confirmation can be a repeat of the same test on a new sample, or a different test that is also abnormal.
  • If two different tests are both above threshold on the same sample, that is sufficient; no repeat is needed.
  • If two tests disagree, the one above threshold should be repeated.
  • Where symptoms are classic and the random glucose is 200 or above, no confirmation is required and treatment starts.
  • Acute illness, infection, steroid treatment and surgery can all raise blood glucose temporarily, so a result taken during these is unreliable for diagnosis.

Why it matters

This step prevents a lifelong label, and sometimes lifelong medication, from resting on a single laboratory value taken on a bad day.

What's still uncertain

  • How often confirmation is actually done in routine practice.
  • How best to classify people whose results sit persistently on the border.

Read more

Next: Which type

Part 3 of 5

Telling type 2 from type 1 and others

Review of many studies

What happens

Not always obvious at presentation.

What the research shows

  • Type 2 is far more common and is the default assumption in an adult with risk factors and gradual onset.
  • Type 1 involves immune destruction of beta cells, tends to come on faster, often with weight loss and ketones, and needs insulin from the start.
  • Adults can develop type 1, and a slowly progressive autoimmune form in adults is frequently misclassified as type 2 at first.
  • Clues that the diagnosis may not be type 2: normal weight, rapid onset, marked weight loss, ketones, a strong personal or family history of autoimmune disease, or failure to respond to oral medicines.
  • Antibody tests and a C-peptide measurement, which reflects the body's own insulin production, can resolve the question when it matters.
  • Other forms exist, including diabetes caused by pancreatic disease, steroid treatment, and rare inherited forms that respond to specific treatments.

Why it matters

Getting the type right changes the treatment entirely. Misclassification is the reason some people spend years on the wrong medicines.

What's still uncertain

  • How often adult-onset type 1 is missed and for how long.
  • When antibody testing is worth doing routinely in adults.

Read more

Next: When A1C misleads

Part 4 of 5

Conditions that distort the result

Review of many studies

What happens

The test measures haemoglobin, so haemoglobin matters.

What the research shows

  • A1C reflects how much glucose has attached to haemoglobin over roughly the preceding two to three months, weighted toward the most recent weeks.
  • Anything that shortens red cell lifespan lowers A1C falsely: haemolysis, recent blood loss, some anaemias, advanced kidney disease, pregnancy and recent transfusion.
  • Anything that lengthens it raises A1C falsely, including iron deficiency and vitamin B12 deficiency before treatment.
  • Some haemoglobin variants, more common in people of African, Mediterranean and South-East Asian descent, interfere with certain A1C assays.
  • Where A1C is unreliable, diagnosis and monitoring should use glucose measurements instead.
  • A1C also does not show variability. Two people with the same A1C can have very different swings between high and low.

Why it matters

A number that looks reassuring can be wrong for a mechanical reason that has nothing to do with blood sugar control.

What's still uncertain

  • How often variant interference is detected in practice.
  • Whether fructosamine or glycated albumin should be used more widely as alternatives.

Read more

Next: What else is checked

Part 5 of 5

The rest of the first visit

Review of many studies

What happens

The diagnosis triggers a wider assessment.

What the research shows

  • Blood pressure, because it is the other main driver of heart and kidney risk and is often raised alongside.
  • A lipid panel, because cholesterol treatment reduces cardiovascular events substantially in diabetes.
  • Kidney function and a urine albumin-to-creatinine ratio. The urine test detects early kidney involvement that blood tests miss, and it is the most commonly skipped test in diabetes care.
  • A dilated eye examination, at diagnosis for type 2 because changes may already be present, then on a schedule.
  • A foot examination including sensation testing and pulses.
  • Liver enzymes, because fatty liver disease is common and frequently unrecognised.
  • Vaccination status, and a discussion of smoking, alcohol and mood.

Why it matters

The first visit is about more than the sugar number. Most of what improves long-term outcomes in diabetes is blood pressure, lipids and screening, not glucose alone.

What's still uncertain

  • The best interval for repeat eye screening in low-risk people.
  • How to improve the low rates of urine albumin testing in routine care.

Read more

More plain-language reading

See something wrong or unclear on this page? Suggest a correction. OwieClear is for education only and isn't medical advice.