A1C Calculator – Convert A1C to Average Blood Sugar (eAG) & Back

Where your A1C sits on the ADA scale

The three ranges doctors use, on one scale. The marker moves live as you type — the shaded band behind it is the range your result falls into.

Keep exploring your health numbers

Three companion tools that pair naturally with blood sugar.

How it works

Three steps from a lab percentage to a number you can picture.

The formula, and the full A1C chart

Understanding your result

What the ranges mean, when A1C can mislead, and why one test wears three different units.

What A1C actually measures

A percentage, a protein, and a three-month window. Understanding the biology explains both the power and the limits of the number.

A1C, also written HbA1c, is the share of hemoglobin — the oxygen-carrying protein in red blood cells — that has glucose attached to it. Glucose in the blood sticks to hemoglobin without any enzyme driving the reaction, a process called glycation, and once attached it stays attached for the life of the red blood cell. A blood sample therefore carries a record of the glucose levels those cells have lived through, not a snapshot of the glucose level at the moment of the draw. That is the fundamental difference between A1C and a finger-prick glucose reading: one is an average over time, the other is a moment.

The window is roughly two to three months because a red blood cell lives about 120 days. Cells are born and retired continuously, so a sample contains cells of every age from new to nearly retired, and the A1C figure is weighted toward the recent past: glucose levels in the last month influence the result more than levels three months ago, because more of the sampled cells were alive during the recent period. A single high-glucose week will move A1C a little, not a lot, and a sustained change takes weeks to show fully, because the older, pre-change cells must retire before the sample reflects the new level. Anyone tracking a change should therefore compare results months apart, not week to week, and should expect the number to move gradually even when daily readings have already changed.

The conversion to average glucose, called eAG, comes from the ADAG study, which measured A1C and continuous glucose together in hundreds of people and fitted the straight line between them: eAG in mg/dL equals 28.7 times A1C minus 46.7. It is a statistical relationship, not a physical law, which means it describes the average person well and any individual person approximately. Two people with the same A1C can have somewhat different true average glucose, and two people with the same true average glucose can have somewhat different A1C, because red cell lifespan and glycation rates vary between individuals. The calculator on this page applies the ADAG line exactly, so its output is the standard estimate. The estimate is the right number to use for general information, and it is not a substitute for a laboratory test interpreted by a clinician.

That distinction — estimate versus diagnosis — runs through everything on this page. A laboratory A1C test, interpreted by a care team alongside other tests and the person’s history, is what supports clinical decisions. A calculator conversion takes a number you enter and returns the ADAG estimate for it. The two can agree closely, and often do, but only the laboratory route carries the quality control, the repeat testing, and the clinical context that a diagnosis requires. Use this page to understand a number you already have, to translate it between unit systems, and to prepare questions for a clinician. Do not use it to decide, on its own, whether a condition is present.

Three languages for one number

Percent, IFCC units, and average glucose all describe the same test. The table below translates between them.

The percentage form, called NGSP percent, is the form used in the United States and the form most American patients see on a lab report. It is a true percentage: the share of hemoglobin carrying glucose. The IFCC form, used across much of the rest of the world, expresses the same quantity in millimoles per mole, and the conversion is IFCC equals A1C minus 2.15, times 10.929. The third form is not a form of A1C at all, but a translation into the units of everyday glucose measurement: estimated average glucose, in mg/dL in the United States or mmol/L elsewhere, where mmol/L equals mg/dL divided by 18.0156. The calculator on this page shows all three at once, because a person comparing a result from one country with a result from another, or a lab report with a glucose meter, needs all three without doing the arithmetic by hand.

Reading the table across a row shows why small percentage changes matter. Moving from 6.0 percent to 7.0 percent is one percentage point, but it corresponds to an average glucose rise from 125.5 to 154.2 mg/dL, a difference of 28.7 mg/dL, which is exactly the slope of the ADAG line. Moving from 7.0 to 8.0 percent adds another 28.7 mg/dL, to 182.9. The relationship is linear by construction, so every percentage point is worth the same 28.7 mg/dL of average glucose, and every tenth of a point is worth 2.87 mg/dL. That linearity is convenient, but it is also a reminder that the conversion is a fitted line: real biology is close to linear over the usual range, and the line is the standard summary of it, not a guarantee for any individual.

The reverse direction works the same way, and the calculator’s second tab exists for it. An average glucose of 180 mg/dL works back to an A1C of 7.9 percent, because 180 plus 46.7, divided by 28.7, is 7.90. An average glucose of 126 mg/dL works back to 6.0 percent, because 126 plus 46.7, divided by 28.7, is 6.02, which rounds to the 6.0 shown in the worked example on this page. Rounding is worth noting: lab reports usually give A1C to one decimal place, and a conversion that lands between two tenths will round to the nearer one. A difference of 0.1 percent between two sources for the same sample is usually rounding, not disagreement, and it corresponds to less than 3 mg/dL of average glucose.

When A1C can mislead, in plain language

The test is reliable for most people, most of the time. The exceptions are worth knowing, because they change what the number means.

Anything that changes how long red blood cells live changes A1C without changing average glucose. Conditions that shorten red cell life, such as some anemias and recent blood loss, leave less time for glucose to attach, so A1C reads lower than the glucose history would suggest. Conditions that lengthen it, such as iron deficiency in some cases, leave more time for attachment, so A1C reads higher. A recent blood transfusion replaces a share of the person’s cells with donor cells carrying a different glycation history, so the result reflects a mixture, not the person’s own glucose. None of these situations makes the test useless, but all of them mean the number should be read alongside other information, and a clinician who knows the history is the right person to do that reading.

Hemoglobin variants add a second kind of exception. A1C measurement assumes the usual adult hemoglobin, and some inherited variants change the molecule enough to interfere with certain laboratory methods, in either direction depending on the method and the variant. The effect is method-specific, which is why a surprising result is sometimes repeated with a different method rather than accepted or dismissed on the first reading. Kidney disease, pregnancy, and some medications can also shift the relationship between A1C and average glucose, for reasons that run through red cell turnover and glycation rates rather than through glucose itself. The pattern across all these cases is the same: A1C is an excellent population tool and a good individual tool, and its exceptions are precisely the situations where individual context matters most.

There is also a subtler limitation that applies to everyone, not just to the exceptions. A1C is an average, and averages hide variation. Two people can share an A1C of 7.0 percent, and therefore an estimated average glucose of 154.2 mg/dL, while one spends most of the day near that average and the other swings between high peaks and low troughs. The average is the same, but the daily experience and the clinical picture are different, which is why continuous glucose monitoring and finger-prick readings remain useful alongside A1C even when A1C is available. The number answers one question — what was the average — and it does not answer the question of how steady the glucose was around it.

The practical conclusion is modest and important. If a result surprises you, or if any of the situations above apply to you, treat the calculator output as a starting point for a conversation, not as a conclusion. Bring the lab report, note any anemia, transfusion, pregnancy, kidney condition, or hemoglobin variant you know of, and ask how the care team weighs A1C against other measurements. That conversation is where a general-information estimate becomes a personal interpretation, and no calculator, including this one, can replace it.

Using an A1C result well, over time

One result is a data point. A series, measured consistently, is a story. Here is how to build the series properly.

First, compare like with like. Use the same laboratory when you can, because methods differ slightly between laboratories, and a change of laboratory can move the result by a tenth or two without any change in glucose. Record the unit system alongside the number, so that a percentage from one report is never compared directly with an IFCC figure from another: 42 mmol/mol and 6.0 percent are the same result in different languages, and comparing them as if they were the same scale is a category error that the conversion table on this page exists to prevent.

Second, space comparisons sensibly. Because A1C reflects roughly the past two to three months, weighted toward the recent month, a repeat measurement a few weeks after a change will show only part of the change, and a repeat the next day will show almost none of it. Clinicians commonly recheck after about three months when tracking a change, which is long enough for the older cells to retire and the new level to dominate the sample. Checking more often is not harmful, but it mostly measures the same cells again, and it can create the misleading impression that nothing is happening when the biology simply has not had time to turn over.

Third, keep the daily picture beside the average. A log of finger-prick readings, or a continuous monitor’s time-in-range summary if one is available, answers the variation question that A1C cannot. If the average is moving in the right direction but the swings are widening, that is worth knowing, and if the average is steady but the lows are increasing, that is worth knowing even more. The two kinds of measurement complement each other: A1C for the long average, daily readings for the shape of each day. Neither alone tells the whole story, and together they give a clinician far more to work with than either does separately.

Fourth, prepare the questions before the appointment, because a result is most useful when it arrives with context. Useful questions include: how does this result compare with my previous results from the same laboratory, is there anything in my history that could make this result read high or low, and what other measurements should sit beside it when we interpret it. Those are questions for a care team, not for a calculator, and they are the reason this page frames everything as general information. The calculator can translate the number into every unit system and place it on the scale. The interpretation, for your body and your history, belongs in a conversation with the people who care for you.

Common unit mistakes, and how to spot them

Most A1C confusion is not about biology. It is about two numbers that look alike and mean different things.

The first confusion is between mmol/L and mmol/mol. An average glucose of 7.0 mmol/L and an A1C of 42 mmol/mol both appear on this page for the same input, and they are easy to swap, because both use millimoles and both sit in the same report. They are different quantities: mmol/L is a concentration of glucose in blood, while mmol/mol is the IFCC expression of the glycated share of hemoglobin. A result quoted as 53 could be an IFCC A1C of 53 mmol/mol, corresponding to 7.0 percent, or it could be nothing at all in glucose units, because glucose in mmol/L rarely reaches 53. If a number seems impossibly large for the unit you think it is in, the unit is probably the other one, and the conversion table on this page will confirm it in seconds.

The second confusion is between a percentage point and a percent change. A move from 6.0 percent to 7.0 percent is a rise of one percentage point, but it is a rise of about 17 percent relative to the starting value, and a rise of 28.7 mg/dL in average glucose. News reports and conversations sometimes describe the first as a small change and the second as a large one, and both descriptions are arithmetically defensible, which is why they cause confusion. When comparing results, always state the unit and the scale: one percentage point on the NGSP scale, or 11 mmol/mol on the IFCC scale, or 28.7 mg/dL of average glucose. With the scale stated, the size of the change is unambiguous.

The third confusion is reading eAG as a measurement rather than an estimate. The letters stand for estimated average glucose, and the estimate comes from the ADAG line fitted to a study population. Your own true average glucose, measured around the clock, will differ from the estimate by some amount, usually small, sometimes larger, depending on individual factors the line cannot capture. Treat eAG as the standard translation of A1C into glucose units, useful for comparing with meter readings and for picturing what the percentage means in everyday terms, and treat a laboratory A1C, interpreted by a care team, as the measurement that carries clinical weight. Keeping those two roles separate prevents the commonest misreading on this page: mistaking a translation for a test result.

Worked examples: reading the conversion both ways

The same formula, run forward from A1C and backward from average glucose.

Start with an A1C of 7.0%. The ADA formula gives eAG = 28.7 × 7.0 − 46.7 = 154.2 mg/dL, which the calculator rounds to 154 mg/dL, or 8.5 mmol/L. In IFCC units the same result is (7.0 − 2.15) × 10.929 = 53 mmol/mol. One lab percentage, three ways of saying it — and the same three numbers appear on the chart above.

Now run it backward. Suppose your meter or CGM average over the last few months is 126 mg/dL. Then A1C = (126 + 46.7) ÷ 28.7 = 6.0%. An average of 183 mg/dL works out to 8.0%, and an average of 212 mg/dL works out to 9.0%. These are estimates from the average, not lab results: the formula describes the average glucose that typically matches an A1C, and an individual person’s average can sit a little above or below it.

The most common reading mistake in this table is direction: 126 mg/dL is an average glucose, not a fasting target and not a single reading. A single fasting value of 126 mg/dL does not by itself give an A1C of 6.0% — the formula needs an average spread over weeks, which is why a meter average and a fasting reading are not interchangeable inputs.

GMI vs A1C: two averages that should not be confused

Your CGM app shows one number, your lab report shows another — and both can be right.

GMI, the Glucose Management Indicator, is an estimate of what your A1C would be, calculated from CGM data. Its formula is GMI (%) = 3.31 + 0.02392 × mean glucose (mg/dL). So a CGM mean of 154 mg/dL gives a GMI of 7.0%, and a mean of 183 mg/dL gives a GMI of 7.7%. The A1C calculator on this page does not compute GMI — it converts a lab A1C to eAG and back — but the two numbers are constantly compared, so the difference is worth understanding.

They differ mainly because they average different windows. A GMI is usually built from 10 to 14 days of CGM readings, while A1C reflects roughly the last two to three months, weighted toward the most recent weeks. Sensor gaps, nights the sensor was off, and days with unusual readings all shift a short CGM window more than they shift a three-month average. On top of that, people glycate hemoglobin at slightly different rates, so the same average glucose can correspond to a slightly different lab A1C from one person to the next.

Notice the last row: the formulas are not identical, so at higher averages GMI and the A1C that the ADAG formula implies can drift apart by a few tenths of a percent. A small gap between your GMI and your lab A1C is therefore expected and is not, on its own, a sign that either number is wrong. A large or persistent gap — or a GMI that keeps disagreeing with your lab result — is worth raising with your clinician, who can look at the CGM data, the lab method and any condition that affects red blood cells. GMI never replaces a lab A1C for diagnosis or for treatment decisions.

Fructosamine: the shorter-window marker

When a two-to-three-week view is more useful than a two-to-three-month one.

Fructosamine measures glycated serum proteins — mostly albumin — rather than glycated hemoglobin. Because albumin turns over much faster than red blood cells, fructosamine reflects average glucose over roughly the last two to three weeks, where A1C reflects roughly the last two to three months. That shorter window is the whole point of the test: it shows recent change sooner, for example after a treatment change, while an A1C is still catching up.

Clinicians sometimes use fructosamine in situations where A1C is known to be unreliable — the same situations listed in the limitations section: anemia, hemoglobin variants, kidney disease and recent transfusion can all distort A1C, because they change red blood cells rather than glucose itself. Fructosamine sidesteps the red blood cell problem, but it brings its own: anything that changes albumin or protein turnover — liver disease, kidney protein loss, thyroid disorders — can distort fructosamine instead. Neither marker is free of caveats, which is why neither is read in isolation.

There is no single reliable formula that converts fructosamine to A1C. Published approximations exist, but they vary by assay and by the population studied, and they carry wide uncertainty, so this page does not present one as a conversion you can rely on. A fructosamine result is best read as its own marker — a shorter-window average, in its own units — and compared with previous fructosamine results from the same lab, rather than translated into an A1C. Fructosamine is not a substitute for A1C, and it is not used on its own to diagnose diabetes. If your A1C and your meter or CGM readings disagree, the right next step is a conversation with your clinician about which marker fits your situation, not a home conversion.

Reading your lab report: percentage, IFCC and eAG side by side

Three numbers for one result — and how to tell which one you are looking at.

A typical US lab report shows A1C as a percentage, the NGSP unit — 6.0%, 6.5%, 7.0%. The same result in IFCC units, used in much of the rest of the world, is a whole number in mmol/mol: 6.0% is 42 mmol/mol, 6.5% is 48 mmol/mol and 7.0% is 53 mmol/mol. Some reports also print eAG, the estimated average glucose, so you can see the percentage and the mg/dL average together. The calculator above reproduces all three from any one of them.

The classic mix-up is reading an IFCC number as a percentage: 42 mmol/mol is not 42%, it is 6.0%. The reverse mistake is entering 42 into a box that expects a percentage and getting a nonsense result — the calculator on this page guards against that by rejecting A1C values above 20% and flagging that a number like 42 or 53 is probably IFCC. A second mix-up is units for glucose itself: an eAG of 7.0 mmol/L is the same as 126 mg/dL, not 7.0 mg/dL. Divide mg/dL by 18.0156 to get mmol/L, or multiply mmol/L by 18.0156 to get mg/dL.

When you compare reports over time, compare like with like — percentage with percentage, or IFCC with IFCC — and from the same lab where possible, since methods differ slightly. A change of 0.5 percentage points, for example from 7.0% to 6.5%, corresponds to an average glucose change of about 14 mg/dL (154 down to 140). That is a real change, but a single result still needs its clinical context: recent illness, medication changes and the conditions in the limitations section can all move a result without glucose itself having changed as much.

A1C units, side by side

Percentage, IFCC, and estimated average glucose: three languages for one number.

Most people meet A1C as a percentage, like 6.0%. Laboratories also report the IFCC value in millimoles per mole, where 6.0% is 42 mmol/mol, converted by the formula IFCC equals 10.93 times the percentage, minus 23.5. A third language translates the percentage back into glucose units: estimated average glucose, or eAG, where 6.0% corresponds to about 125.5 mg/dL, or 7.0 mmol/L, by the ADA formula of 28.7 times the percentage, minus 46.7. The conversion chart on this page shows all three side by side, so a result in any language can be read in the other two without arithmetic.

Two related tools sit behind the most common searches. An average A1C calculator runs the conversion in reverse, turning an average glucose reading from a meter or a continuous glucose monitor into an estimated A1C, using the same ADA formula backwards. An A1C conversion chart, like the one in the formula section above, lists percentages against their eAG and IFCC equivalents. Both are estimates built on population averages: they translate between scales, they do not measure your glucose, and they cannot see the factors, such as anaemia or kidney disease, that can move a reading away from what average glucose alone would predict. The companion guide on what affects an A1C reading explains those factors in general terms.

Related guides

Deeper explanations of the conversions on this page.

More health tools

Weight, energy and body composition — all free, all instant.