You can estimate your A1c from average glucose readings using a straightforward formula: divide your average glucose in mg/dL by 28.7, then add 1.63. For example, an average glucose of 154 mg/dL works out to an estimated A1c of about 7%. The math is simple enough to do on a phone calculator, and the result will land close to what a lab test would show for most people.
The Core Formula
The most widely used conversion comes from a large international study that compared continuous glucose monitoring data against lab-measured A1c in hundreds of participants. The original equation is written as:
Estimated average glucose (mg/dL) = (28.7 × A1c) − 46.7
Since you’re going the other direction, solving for A1c from glucose, just rearrange it:
A1c (%) = (average glucose in mg/dL + 46.7) ÷ 28.7
If you measure glucose in mmol/L (common outside the U.S.), the rearranged formula is:
A1c (%) = (average glucose in mmol/L + 2.59) ÷ 1.59
The correlation between average glucose and lab A1c is strong (r² = 0.84), but not perfect. About 16% of the variation comes from individual biological differences, which means two people with identical average glucose levels can have slightly different lab A1c results.
Quick Reference Table
If you’d rather skip the math, this table maps common average glucose values to their estimated A1c:
- 100 mg/dL (5.6 mmol/L): ~5.1%
- 120 mg/dL (6.7 mmol/L): ~5.8%
- 140 mg/dL (7.8 mmol/L): ~6.5%
- 154 mg/dL (8.6 mmol/L): ~7.0%
- 183 mg/dL (10.2 mmol/L): ~8.0%
- 212 mg/dL (11.8 mmol/L): ~9.0%
- 240 mg/dL (13.3 mmol/L): ~10.0%
- 269 mg/dL (14.9 mmol/L): ~11.0%
Each full percentage point of A1c corresponds to roughly a 29 mg/dL change in average glucose. That’s a useful mental shortcut: if your average drops by about 30 mg/dL, expect your A1c to fall by roughly one point.
CGM Users: The GMI Formula
If you wear a continuous glucose monitor, your reports likely show a number called the Glucose Management Indicator (GMI) rather than an estimated A1c. GMI uses a slightly different formula developed specifically from CGM data:
GMI (%) = 3.31 + (0.02392 × mean glucose in mg/dL)
For a mean sensor glucose of 154 mg/dL, the GMI comes out to about 7.0%, which happens to match the standard formula at that point. But the two formulas can diverge at higher or lower glucose levels because they were derived from different study populations.
Your GMI and your lab A1c won’t always match. The difference isn’t an error. It reflects real biological variation in how quickly your red blood cells accumulate glucose or how long those cells survive. Some people consistently run a lab A1c that’s 0.5% higher or lower than their GMI predicts. If you notice a persistent gap, your GMI still tracks trends in your glucose control accurately, even if the absolute number doesn’t line up with your lab result.
How to Calculate Your Average Glucose
The formula is only as good as the average you feed into it. A single fasting reading won’t cut it, because A1c reflects glucose levels around the clock, including the spikes after meals and the dips overnight.
If you’re using a fingerstick meter, you’ll get the most representative average by testing at varied times: before and after meals, at bedtime, and occasionally overnight. Aim for at least four to seven readings per day spread across different times for at least two to three weeks. Then average all those readings together. Checking only first thing in the morning will underestimate your true average, since post-meal highs won’t be captured.
CGM makes this much easier. The sensor captures a reading every few minutes, generating hundreds of data points per day. A 14-day CGM average is a solid basis for the GMI calculation. Most CGM apps display this average and the GMI automatically.
International Units (mmol/mol)
Outside the United States, A1c is often reported in mmol/mol (the IFCC standard) rather than a percentage. If your glucose data gives you an A1c in percent and you need mmol/mol, the conversion points are:
- 5.0% = 31 mmol/mol
- 6.0% = 42 mmol/mol
- 7.0% = 53 mmol/mol
- 8.0% = 64 mmol/mol
- 9.0% = 75 mmol/mol
- 10.0% = 86 mmol/mol
Each 1% increase in the NGSP percentage equals roughly 11 mmol/mol.
When the Calculation Won’t Be Accurate
The glucose-to-A1c formula assumes your red blood cells have a normal lifespan of about 120 days. Several conditions change that assumption and throw off the calculation in predictable directions.
Conditions that shorten red blood cell life, like hemolytic anemia, recovery from significant blood loss, or being on dialysis, will make a lab A1c read falsely low. Your calculated estimate from glucose readings would actually be more accurate than the lab test in these situations, because glucose data isn’t affected by red blood cell turnover.
Iron deficiency anemia does the opposite. It’s associated with higher A1c readings than glucose levels would predict. This is especially relevant in late pregnancy, where iron deficiency can push lab A1c upward even in people without diabetes.
Inherited hemoglobin variants (such as sickle cell trait or hemoglobin C trait) can interfere with A1c measurement in the lab, making results unreliable regardless of the assay. If you carry one of these variants, glucose-based monitoring becomes the primary way to track long-term control. Kidney failure also complicates A1c accuracy: chemically modified hemoglobin in people with renal disease can skew lab results, and studies suggest A1c underestimates glucose levels in dialysis patients.
Even without any of these conditions, individual biology creates a margin of error. Some people glycate hemoglobin faster or slower than average. If your calculated A1c from glucose data consistently disagrees with your lab result by more than 0.5%, the pattern is likely real and worth noting when you and your care team interpret results.