Average Glucose to A1C Estimator
Estimate the A1C range that corresponds to your average blood glucose, from a meter or CGM average in mg/dL or mmol/L. The answer is deliberately a range, not a single percentage — this direction of the calculation has built-in uncertainty that one number would hide.
Estimate
Enter your average glucose
Type the average from your meter or CGM and pick its unit; the estimated A1C range updates as you type.
Estimated A1C range
Enter an average glucose above to see the estimated A1C range.
The math
How the estimate is calculated — and why it is approximate
A1C (%) ≈ (eAG + 46.7) ÷ 28.7
This equation is an algebraic rearrangement of the regression the ADAG study (Nathan et al., Diabetes Care 2008) published for the opposite direction — turning an A1C percentage into an estimated average glucose. The study validated that forward direction; running it backwards, as this page does, is a convenience the study itself never named or certified. Regression lines are asymmetric: the line fitted to predict glucose from A1C is not the line you would get fitting A1C from glucose, so the backwards estimate carries extra error, and that error grows near clinically important cut-offs.
Why you get a range, not a number
The scatter around the ADAG regression is measurable: the reported standard deviation is about 15.7 mg/dL of glucose. Propagating that scatter through the slope of 28.7 gives 15.7 ÷ 28.7 ≈ 0.55 percentage points of A1C — roughly ±0.5% around the central value. The tool keeps that half-width unrounded, adds and subtracts it from the center, and only then rounds each endpoint to 0.1%. Two people with the same true average glucose can genuinely sit at opposite ends of the range shown.
This page versus the A1C to eAG calculator
The two tools are mirror images with different standing. The A1C to eAG calculator applies the published ADAG regression in its original, validated direction and can therefore report a value. This estimator reverses the same relationship to answer a different practical question — “what A1C would this average likely map to?” — and honestly reports the uncertainty of that reversal as a range. Neither tool measures anything: for a real A1C you need an NGSP-certified laboratory test.
FAQ
Average glucose and A1C questions
Why does this estimator show a range instead of one A1C number?
Because the underlying data cannot support a single number in this direction. The ADAG regression that links A1C and average glucose has real scatter around it — the reported standard deviation is about 15.7 mg/dL — and when you divide that scatter by the slope of 28.7, it propagates into roughly ±0.5 percentage points of A1C. Printing one value would fake a precision the study never measured, so this page computes the center, applies the half-width of 15.7 ÷ 28.7 without pre-rounding, and only then rounds each endpoint to 0.1%.
How is this different from the A1C to eAG calculator?
The two pages run in opposite directions. The A1C to eAG calculator applies the regression exactly as the ADAG study published it: A1C in, estimated average glucose out. This page rearranges that equation algebraically to run backwards, from average glucose to A1C — a direction the study never validated as its own equation. Regression lines are not symmetric, so the backwards estimate inherits extra uncertainty, which grows near clinically important cut-offs. That is also why this page reports a range while the forward page can report a value.
My lab A1C is different from this estimate. Which one should I trust?
The laboratory result. This estimator starts from a meter or CGM average, which covers a different time window than A1C, weights every reading equally where A1C weights recent weeks more heavily, and misses readings you never took. On top of that, red blood cell lifespan differs between people, and pregnancy, hemoglobin variants, anemia, and kidney disease can all move A1C independently of average glucose. Treat the range here as orientation for understanding your numbers, and treat the NGSP-certified laboratory value as the real measurement.