Anion Gap Normal Range by Age & Lab Method
The anion gap normal range is not a single number. It varies by patient age, by the laboratory's measurement method, and by whether potassium is included in the formula. This reference table consolidates every accepted reference interval in one place.
The reference range, at a glance
The anion gap normal range most clinicians memorise is 8–12 mEq/L. That figure, derived from the conventional formula Na⁺ − (Cl⁻ + HCO₃⁻), dates from the era of colorimetric and flame-photometry electrolyte measurement. Modern laboratories using ion-selective electrodes (ISE autoanalyzers) report slightly higher chloride values, which compresses the calculated gap into a tighter 3–11 mEq/L interval (Kraut & Madias, StatPearls 'Serum Anion Gap'; LITFL Acid-Base). Both are "correct" — the right answer is the one printed on your laboratory's report. You can check any result against either convention in the anion gap calculator, which flags high, low, and normal against the locally relevant band.
Comprehensive range table by age & method
| Population | Conventional method (mEq/L) | ISE autoanalyzer (mEq/L) | Notes |
|---|---|---|---|
| Neonate (0–28 days) | 5–15 | 5–13 | Wider interval; albumin and organic anions vary widely in early life. |
| Infant (1–12 months) | 7–16 | 5–14 | Slightly elevated versus adult baseline; lactate and ketone flux common. |
| Child (1–18 years) | 8–14 | 3–12 | Approaches adult range through childhood; see pediatric specifics below. |
| Adult (18–65 years) | 8–12 | 3–11 | The canonical reference interval used in most clinical decision tools. |
| Elderly (> 65 years) | 7–11 | 2–10 | Mildly lower baseline, driven by age-related fall in serum albumin. |
| Pregnancy (3rd trimester) | 7–10 | 2–9 | Physiologic hemodilution and respiratory alkalosis lower the expected gap. |
Pediatric anion gap — why children differ
Infants and young children run a slightly higher pediatric anion gap than adults, principally because of higher baseline lactate and ketone production and a developing renal acidification capacity. Neonates in particular have a wide reference interval (5–15 mEq/L), so a gap that would be frankly elevated in an adult can sit within the normal band for a newborn. When interpreting a pediatric result, always use the age-appropriate range from the reporting laboratory rather than the adult 8–12 value.
The potassium-inclusive range
When potassium is folded into the formula — (Na⁺ + K⁺) − (Cl⁻ + HCO₃⁻) — the entire reference interval shifts upward because K⁺ (~4 mEq/L) is added to the cation side. The potassium-inclusive normal anion gap is therefore roughly 10–20 mEq/L. Most modern practice omits potassium for simplicity (it is small and tightly regulated), reserving the K-inclusive equation for hyperkalemic acidosis and renal physiology teaching. Mixing the two conventions is a common source of error: never compare a K-inclusive result against the 8–12 standard without adjusting.
Albumin-adjusted expected normal
Albumin contributes roughly 75% of the "normal" anion gap, so when it falls the expected normal falls with it. Each 1 g/dL decrease in albumin lowers the expected anion gap by about 2.5 mEq/L (Figge et al., Critical Care Medicine). The albumin-corrected formula — expected AG = observed AG + 2.5 × (4.0 − albumin g/dL) — restores the comparison to a normal-albumin baseline.
In a hypoalbuminemic ICU patient (albumin 2.0 g/dL), the uncorrected "normal" gap is not 8–12 but closer to 3–7 mEq/L. A raw gap of 10 — which looks reassuringly mid-range — is in fact already elevated for that patient and may signal an early high-anion-gap metabolic acidosis that the uncorrected number hides. Always apply the albumin correction before declaring a gap normal in any critically ill or hypoalbuminemic patient.
Why laboratories differ
The single biggest source of inter-laboratory disagreement is chloride measurement. ISE autoanalyzers report Cl⁻ values roughly 2–4 mEq/L higher than older methods, which directly shrinks the calculated anion gap. Sodium and bicarbonate assays are more consistent across platforms. The practical consequence is that a patient serially monitored across two hospitals can appear to have a "dropping" anion gap purely from method differences. For a worked discussion of this variation and the formula-level implications, see the normal anion gap range derivation page. When in doubt, interpret each result against the reference interval supplied by the reporting lab — never a universal number.
For interpretation of specific anion gap values (14, 15, 19, 20 mEq/L and above), see our specific anion gap value guide.