Anion gap = sodium − (chloride + bicarbonate).
For decades the textbook anion gap range was 8 to 16 mEq/L. In 1990, a laboratory study using newer electrode analyzers found the range had shifted down, to 3 to 11 in one of their labs, mainly because measured chloride ran higher. Most healthy people tested on that equipment had a gap of 6 or less.
Many modern labs print a range of about 3 to 9 mEq/L, but your own lab's printed range is the only one that fits your result. Reviews also describe real variation between healthy individuals, so a value just outside the range is common and means little on its own.
A high gap is most often associated with metabolic acidosis, a buildup of acids the panel does not measure directly. A low gap most often reflects a lab error or low albumin. Neither direction names a cause by itself, and both are read with the rest of the panel.
Albumin carries much of the negative charge the gap stands in for. In a study of intensive care patients, each 1 g/dL drop in albumin made the observed gap read about 2.5 mEq/L lower. Low albumin can therefore hide a raised gap, which is why some clinicians adjust for it.
Beyond acid buildup, a high gap can reflect lab error, metabolic alkalosis or a high phosphate level. A low gap can also be associated with abnormal antibody proteins called paraproteins, or with high blood levels of lithium, bromide or iodide.
Sample handling matters too. Bicarbonate can drift down when a tube is underfilled or left open before analysis, which raises the calculated gap without any change in the body. A long fast or very hard exercise can raise acids such as ketones or lactate for a while and nudge the gap up.
Some medicines shift the inputs as well. Diuretics and steroid medicines can alter bicarbonate, and very high lithium levels can lower the gap, as reviews of the anion gap describe. None of this is a reason to change a medicine without talking to the prescriber.
A gap a point or two outside the range with a normal bicarbonate and no symptoms is often simply repeated at the next routine panel. Because each lab sets its own range for its own analyzer, a retest is most useful at the same lab.
A high gap with a low bicarbonate, or alongside vomiting, fast breathing, confusion or very high blood sugar, is worth raising with a clinician promptly. A low gap that persists on repeat testing is also worth discussing, since reviewers describe it as an overlooked clue that can lead to further protein testing.
Before a repeat, it helps to note anything unusual around the first draw, such as a long fast, a hard workout, a recent illness or a new medicine. Those details often explain a one off shift and can save a round of extra testing.
The gap is calculated from three separately measured electrolytes, so the error in each one adds up. A small calibration change in chloride alone can move the result by several points, which is part of why the reference range dropped as analyzers changed.
Clinicians have compared the change in the gap with the change in bicarbonate to look for mixed acid and base problems. A major review notes that the relationship between the two varies more than once assumed, so it cannot be relied on alone. The gap is a screening clue that needs blood gases and the clinical picture behind it.
This calculator leaves potassium out, as many labs do. Some labs add it, which raises the result by whatever the potassium value is. When you compare results over time, check that each one was calculated the same way, with or without potassium, and ideally on the same lab's analyzer.
Take a sodium of 140, a chloride of 104 and a bicarbonate, often printed as CO2, of 24, all in mEq/L. The calculator adds chloride and bicarbonate to get 128, then subtracts that from sodium. 140 minus 128 gives an anion gap of 12 mEq/L.
Against the old 8 to 16 range, 12 sits in the middle. Against a modern range of about 3 to 9 it reads high, which shows why the printed range matters. If the same person's albumin were 1 g/dL below their lab's normal, the albumin adjusted gap would be about 14.5. A clinician would read this with bicarbonate, glucose and kidney markers.
Blood gas testing measures pH and carbon dioxide directly, so it shows whether acid and base balance is actually off, which the gap alone cannot do. The gap then helps sort what kind of change might be present. In a hospital, clinicians often look at the gap and a blood gas together, along with glucose, kidney markers and lactate, before deciding what a change in either one means.
The albumin adjusted gap is a refinement of the same formula, adding about 2.5 for each 1 g/dL that albumin sits below normal. It earns its keep when albumin is low, as in many hospital patients. Other approaches model every charged particle in the blood, but they need more measurements than a standard panel provides.
The urine anion gap and the osmolal gap answer different questions, about how the kidneys handle acid and about unmeasured small molecules in blood. Neither is interchangeable with the serum anion gap this calculator returns. If a clinician mentions one of them, it is usually because the serum gap or bicarbonate raised a specific question those tests are better placed to answer.
The Metabolic and insulin panel measures the values this calculator uses. Related: sodium, chloride, co2.
An anion gap calculator takes three numbers from a standard metabolic panel, sodium, chloride and bicarbonate, and applies one subtraction, sodium minus the sum of chloride and bicarbonate. The result, in mmol/L, estimates the unmeasured anions in your blood and gives a clinician a clue about acid base balance. It is not a diagnosis.
The calculator above does the arithmetic from your own values. What follows explains where each input comes from, why two laboratories can print different normal ranges for the same formula, what a high or low figure tends to point toward, and when a result is worth raising with a clinician promptly.
The anion gap is sodium minus the sum of chloride and bicarbonate, with all three values in mmol/L (numerically the same as mEq/L).
Blood is electrically neutral, so the positive and negative charges in plasma balance exactly. Routine chemistry measures only a handful of them, mainly sodium among the positively charged ions and chloride and bicarbonate among the negatively charged ones. Several negatively charged substances are never measured directly, including albumin, phosphate, sulfate and organic acids such as lactate and ketoacids. Subtracting what is measured on the negative side from what is measured on the positive side leaves a difference that stands in for those unmeasured anions, and albumin normally accounts for the largest single share of it 1.
All three inputs come from a basic or comprehensive metabolic panel, the standard chemistry drawn in most routine blood work. Bicarbonate is often printed as CO2 or total CO2 on a report, which is the same figure the formula wants. A worked example makes the arithmetic obvious. Sodium 140, chloride 104 and bicarbonate 24 gives 140 minus 128, so the gap is 12 mmol/L.
One detail changes the answer. Some laboratories and some online tools add potassium to the positive side, giving sodium plus potassium minus the sum of chloride and bicarbonate. Because potassium usually sits at around 4 mmol/L, that version returns a number several points higher and carries its own, higher reference interval. Comparing a result produced by one version against the normal range for the other is one of the most common ways people confuse themselves. The calculator on this page uses the three input form.
Read your figure against the interval printed on your own lab report, then read it alongside the rest of the panel, because the same number carries different weight depending on what surrounds it.
Inside your laboratory's interval. A normal gap is reassuring, but on its own it does not rule out an acid base disturbance. That is why the bicarbonate value itself, not only the gap, belongs in the reading.
Below the interval. A low gap is uncommon and is usually explained by something other than acid chemistry. Research suggests that low values most often reflect a reduced concentration of negatively charged protein, most commonly albumin, or laboratory interference, rather than a primary acid base problem 1. Abnormal proteins and a few unusual ingestions can also compress it. A single low value with a normal albumin is generally worth repeating before anyone interprets it.
Modestly above the interval. A mild elevation is frequently non specific. Dehydration, a recent bout of hard exercise that transiently raises lactate, reduced kidney function and ordinary day to day laboratory variation can each nudge the figure up. The practical next step is a repeat panel with albumin, bicarbonate and kidney markers drawn in the same session, rather than a conclusion from one snapshot.
Well above the interval. A clearly raised gap tells a clinician that unmeasured acid anions are accumulating, and it narrows the search to broad categories, including ketoacids, lactate, anions retained when the kidneys are failing, and certain ingestions 2. It does not identify which of those applies; history, symptoms and further testing do that. Studies suggest that a raised anion gap is associated with worse clinical outcomes across several hospital populations, which is part of why a high value is treated as a reason to look harder rather than to wait and see 3.
A markedly raised or markedly low result is suboptimal and worth discussing with a clinician promptly. An abnormal result alongside vomiting, fast or deep breathing, drowsiness or confusion needs urgent medical care the same day.
Reference intervals differ because measurement methods differ, not because the underlying biology changes.
Chloride and bicarbonate are measured on analyzers whose calibration varies between instruments and manufacturers, and a small systematic difference in either of those two numbers shifts the calculated gap directly. Many laboratories using current ion selective electrodes report typical intervals somewhere around 3 to 11 mmol/L. Typical lab reference ranges also vary by laboratory, age and sex, so the interval printed next to your own result is the one that applies to it.
This is why a value should not be judged against a range found online, and why the trend in your own results, measured by the same laboratory over time, usually tells you more than one figure compared against a generic band.
The arithmetic is exact, the inputs are not, and a handful of predictable factors make the gap read lower or higher than the underlying chemistry would suggest.
Albumin is the large one. Because albumin contributes the biggest share of the unmeasured anions, a low albumin level lowers the calculated gap, and a meaningful acid load can sit hidden behind a figure that looks ordinary. Clinicians routinely adjust the result upward when albumin is low before interpreting it, which is the main reason albumin belongs on the same draw as the electrolytes 1.
Other sources of error matter too. Sample handling counts, because bicarbonate can drift if a tube stands or is exposed to air before analysis. Very high lipid or protein concentrations can distort sodium measurement on some analyzers, which moves the gap with it. Including or excluding potassium changes the figure, as described above. A mixed acid base picture can partly cancel itself out, leaving a normal looking gap while two opposing disturbances are present 5. And one reading captures a single moment, while real values move with hydration, food, illness and training load.
None of this makes the calculation pointless. It makes it a screening step. Research suggests that the anion gap may help a clinician narrow the list of possible causes once metabolic acidosis has been identified, which is a different job from establishing what is actually wrong 4.
The gap is a derived figure, not a measurement of any single substance, so it can point in a direction without naming a destination.
A raised value says that something negatively charged and not routinely measured has accumulated. Several very different processes produce that same signal, which is why the clinical approach starts from the gap and then works through history, examination and targeted testing to separate them 2. A normal value, equally, is compatible with an acid base disturbance that keeps chloride and bicarbonate moving in step. The honest reading of any single result is observation first, then possible meaning, then a conversation with someone who can see the rest of your picture.
Associations between the anion gap and clinical outcomes are described in populations 3, and a population association does not forecast what will happen to one person with one result on one day.
The gap is close to meaningless in isolation, so read it with the panel that produced it.
Bicarbonate is the first companion value, because it indicates whether an acid base disturbance is present at all. Albumin is the second, because it calibrates how much the gap can be trusted. Sodium and chloride deserve attention in their own right, particularly the chloride trend, which often moves in the opposite direction to the gap. Kidney markers such as creatinine and estimated filtration rate add context, since retained anions in reduced kidney function are one recognised reason for a higher figure. In advanced chronic kidney disease, a higher time updated anion gap is associated with greater risk of progression to kidney failure in one cohort study, although that is a group level association rather than an individual prediction 6.
Glucose, and when a clinician orders them lactate or a blood gas, complete the picture in a clinical setting. If you are following these values over time rather than during an acute illness, the useful habit is to watch the direction of travel across repeat panels instead of one decimal place on one report.
Sodium, chloride and bicarbonate sit on the standard metabolic chemistry included in most routine blood panels, so most people already have everything the formula needs on a report they have been handed.
FitScript Essential is the baseline panel in a labs first membership that covers 150+ biomarkers tested per year, with results interpreted by Atlas, FitScript's AI health advisor, and any treatment decision made by a licensed provider after reviewing the member. The practical thing to confirm on your own report is whether albumin and kidney markers were drawn in the same session as the electrolytes, because those are the values that make a calculated gap interpretable.
Any result outside your laboratory's interval is worth raising with a clinician, and the further outside it sits, the sooner that conversation should happen.
Seek urgent care the same day if an abnormal gap comes with vomiting that will not settle, breathing that is unusually fast or deep, abdominal pain, drowsiness, confusion, or a known reason for acid build up such as poorly controlled diabetes. Those features change the question from what does my number mean to how quickly do I need to be seen.
For a figure that is mildly outside the range with no symptoms, the usual path is a repeat panel including albumin, bicarbonate and kidney markers, plus a review of medications, hydration and any recent illness. Management of an acid base disturbance belongs to the clinician who can examine you and see the whole picture, and no medication or supplement decision should ever be made from a calculator output. Never set an abnormal value aside because an online tool made it look minor.
This page is educational and is not medical advice. Discuss your results, and any treatment, with a clinician or licensed provider who can review your full history.
For what the marker measures on its own and how laboratories report it, see Anion Gap. For the baseline chemistry that supplies sodium, chloride and bicarbonate, see FitScript Essential.
Sodium minus the sum of chloride and bicarbonate. Some labs include potassium with sodium.
Typical values shifted lower as laboratory analyzer methods changed, so each lab sets a range for its own equipment.
On a metabolic panel, total CO2 is mostly bicarbonate and is the value used in the calculation.
There is no single universal figure. Many laboratories using current analyzers report typical intervals around 3 to 11 mmol/L. Typical reference ranges vary by laboratory, age and sex, so use the interval printed on your own report.
A raised gap suggests that unmeasured acid anions are accumulating, which points a clinician toward broad categories such as ketoacids, lactate, anions retained in kidney failure, or certain ingestions. It identifies a direction to investigate, not a condition, so it always needs clinical interpretation alongside symptoms and other tests.
Yes, and a low value is uncommon. It most often reflects a reduced level of negatively charged protein, usually albumin, or laboratory interference, rather than an acid base problem. A single low result with a normal albumin is generally worth repeating before anyone draws conclusions from it.
Albumin is the largest contributor to the unmeasured anions the gap stands in for. When albumin is low, the calculated gap falls, and a real acid load can hide behind a normal looking number. Clinicians adjust for low albumin before interpreting the figure, which is why both are best drawn together.
Both forms are used. Including potassium raises the result by roughly the potassium concentration, usually around 4 mmol/L, so it requires a higher reference interval. The important thing is consistency, comparing a result against the normal range for the same version of the formula that produced it.
They can. Dehydration shifts electrolyte concentrations, and a hard training session can transiently raise lactate, which is one of the unmeasured anions. Mild, temporary elevations are common for these reasons, which is why a repeat panel under ordinary conditions is more informative than a single post workout draw.
It usually means the result should be reviewed in context rather than acted on alone. A clinician may repeat the panel with albumin, bicarbonate and kidney markers, review medications and history, and order further tests if warranted. Any result outside your laboratory's interval is worth discussing with a clinician.