A1C to Average Glucose Calculator
Convert between HbA1c and estimated average glucose (eAG) in both mg/dL and mmol/L, using the ADAG study formula.
How to use this calculator
- 1Choose which direction you want to convert.
- 2Enter the A1C percentage, or an average glucose reading in mg/dL or mmol/L.
- 3Both glucose units are shown in the result, since labs in different countries report different ones.
How the calculation works
eAG (mg/dL) = 28.7 × A1C − 46.7 A1C = (eAG + 46.7) ÷ 28.7 mmol/L = mg/dL ÷ 18.0182- A1C
- Glycated haemoglobin, as a percentage
- eAG
- Estimated average glucose over roughly the preceding 2–3 months
The formula comes from the A1c-Derived Average Glucose (ADAG) study, which compared HbA1c against continuous glucose monitoring in 507 participants across ten international centres. The regression had an r² of 0.84 — a strong relationship, but not a perfect one.
That r² is the honest caveat: the formula describes the average relationship across a population, so an individual's true average glucose can differ from their eAG. It is a translation aid, not a measurement.
The mg/dL to mmol/L factor of 18.0182 comes from glucose's molar mass (180.156 g/mol) and is exact for the unit conversion itself, unlike the A1C relationship which is statistical.
Worked example
An HbA1c of 7%
- 1.eAG = 28.7 × 7 − 46.7 = 200.9 − 46.7 = 154.2 mg/dL.
- 2.Converting to mmol/L: 154.2 ÷ 18.0182 = 8.56 mmol/L.
- 3.An A1C of 7% is above the 6.5% diabetes threshold, and is a common treatment target for many adults with diabetes.
Result: 154 mg/dL, or 8.6 mmol/L
What HbA1c actually measures
Glucose in the bloodstream attaches to haemoglobin in red blood cells, a process called glycation, and once attached it stays for the life of that cell. HbA1c measures what proportion of haemoglobin has been glycated, which makes it a running average of blood glucose over roughly the preceding two to three months — the typical lifespan of a red blood cell being about 120 days.
That is fundamentally different from a fingerstick reading, which captures a single moment. A person can have a normal reading before breakfast and a substantially raised A1C, because the A1C is integrating everything that happened between tests, including overnight and post-meal excursions no spot check saw.
Why eAG exists at all
A1C is reported as a percentage, which is a scale nobody encounters anywhere else in daily glucose management. Someone checking their glucose several times a day thinks in mg/dL or mmol/L, and "7%" does not connect intuitively to those numbers.
Estimated average glucose was introduced to bridge that gap: it restates the same A1C result in the units patients already use. An A1C of 7% becomes "about 154 mg/dL on average", which is directly comparable to the readings on a meter.
Where the estimate breaks down
Several conditions make A1C an unreliable proxy for average glucose, and the formula cannot detect any of them.
- Anything affecting red-cell lifespan — haemolytic anaemia, recent blood loss or transfusion, and some medications shorten cell life, giving cells less time to glycate and producing a falsely low A1C.
- Iron-deficiency anaemia — tends to push A1C the other way, producing a falsely raised result.
- Haemoglobin variants — sickle cell trait, thalassaemia and other variants interfere with some A1C assay methods, though not all.
- Pregnancy and kidney disease — both alter red-cell turnover enough to make A1C less reliable as an average-glucose proxy.
- Individual glycation rate — even in healthy people, some glycate haemoglobin faster than others at the same glucose level — the ADAG r² of 0.84 is exactly this variation showing up in the data.
A1C and time in range
A1C compresses months into one number, which is its strength and its blind spot. It cannot distinguish someone with steady glucose near their average from someone swinging between hypoglycaemia and very high readings that happen to average out the same. Both produce the same A1C, and the second carries meaningfully more risk.
This is why continuous glucose monitoring introduced "time in range" as a complementary measure — the share of the day spent within target. A1C remains the long-established outcome measure that clinical trials and treatment targets are built around, but it is increasingly read alongside variability rather than alone.
What this assumes, and where it stops
Assumptions
- Red blood cell lifespan and glycation behave typically. Several common conditions break this — see the limitations.
- The ADAG regression applies, which was derived from people with type 1 diabetes, type 2 diabetes and without diabetes across ten international centres.
Limitations
- eAG is a population estimate with an r² of 0.84, not a measurement. An individual's true average glucose can differ meaningfully from the figure shown.
- Unreliable where red-cell lifespan is altered — anaemia, recent transfusion, haemoglobin variants, pregnancy and kidney disease all distort A1C.
- Cannot show glucose variability. Steady glucose and wide swings can produce an identical A1C despite carrying different risk.
- Diagnostic bands are shown for context only. Diagnosis is a clinical decision made on repeat testing, not on a converted figure.
Common questions
What does an A1C of 7% mean in blood sugar terms?
About 154 mg/dL, or 8.6 mmol/L, as an average over the preceding two to three months. Seven percent is a common treatment target for many adults with diabetes, though targets are individualised — tighter for some, looser for others depending on age, duration of diabetes and hypoglycaemia risk.
Why does my meter average not match my A1C?
Several reasons. Fingerstick readings sample particular moments and usually cluster around routine times, so they under-sample overnight and post-meal periods. A1C also weights recent weeks more heavily. And individual glycation rates vary — the ADAG relationship holds across a population but not exactly for any one person.
Can I use this to diagnose diabetes?
No. The bands are shown for context, but diagnosis requires clinical assessment and typically repeat testing, and several conditions make A1C unreliable as a glucose proxy. This converter restates a number you already have — it is not a diagnostic tool.
Formula and content last reviewed on .
Results are estimates for information only, not professional advice.
Related calculators
Tools people commonly use alongside the a1c to average glucose calculator.