Reference

Electrolyte Panel & BMP Reference Ranges

The basic metabolic panel (BMP) is the single test that feeds the anion gap. Here are the reference ranges for every analyte on the panel, what each one measures, and how the four electrolytes combine to produce the anion gap.

The basic metabolic panel

The basic metabolic panel (BMP) is a standard eight-analyte blood test that screens renal function, electrolyte balance, and glucose homeostasis. Four of its eight components — sodium, chloride, bicarbonate (reported as total CO₂), and potassium — directly determine the anion gap (Kraut & Madias, StatPearls 'Serum Anion Gap'). The remaining four (BUN, creatinine, glucose, calcium) provide the clinical context needed to interpret a high or normal gap. Every value quoted below is an adult conventional-unit reference interval; verify against your reporting laboratory's own range, and consult the master normal ranges table for age- and method-specific values.

Reference range table

Analyte Symbol Adult reference range Unit Feeds the AG?
Sodium Na⁺ 135–145 mEq/L Yes — major measured cation
Potassium K⁺ 3.5–5.0 mEq/L Optional (K-inclusive variant)
Chloride Cl⁻ 98–107 mEq/L Yes — major measured anion
Bicarbonate / Total CO₂ HCO₃⁻ 22–26 mEq/L Yes — major measured anion
Blood urea nitrogen BUN 7–20 mg/dL No — context (renal function)
Creatinine Cr 0.6–1.2 mg/dL No — context (renal function)
Glucose Glu 70–99 (fasting) mg/dL No — context (DKA screen)
Calcium (total) Ca²⁺ 8.5–10.5 mg/dL No — minor unmeasured cation

What each analyte measures

Sodium (Na⁺) is the dominant extracellular cation and the principal determinant of serum osmolality. It is the cation side of the anion gap equation, so a low sodium (e.g. from dilution with hypo-osmolar fluids) can artefactually narrow the calculated gap, while hypernatremia can widen it. Potassium (K⁺) is included only in the potassium-inclusive variant of the formula; because it is small relative to sodium, most clinical practice omits it. Chloride (Cl⁻) and bicarbonate (HCO₃⁻) are the two measured anions whose sum is subtracted from sodium. A falling bicarbonate is the hallmark of metabolic acidosis; a rising chloride compensates and narrows the gap (hyperchloremic acidosis), while a stable chloride leaves the gap to widen (high anion gap acidosis).

BUN and creatinine are not inputs to the formula, but they flag the renal-failure limb of HAGMA — accumulated uremic anions raise the gap in advanced CKD or AKI. Glucose screens for diabetic ketoacidosis: a markedly elevated glucose with ketosis drives the high anion gap through β-hydroxybutyrate accumulation. Calcium is a minor unmeasured cation; clinically relevant only when markedly deranged or in the Stewart strong-ion framework.

How the panel feeds the anion gap

Plug the four electrolyte values into the formula Na⁺ − (Cl⁻ + HCO₃⁻) and you have the anion gap. A worked example from a normal panel: Na⁺ 140, Cl⁻ 104, HCO₃⁻ 24 → 140 − (104 + 24) = 12 mEq/L, the upper bound of the conventional normal range. The anion gap calculator performs this arithmetic, flags the result against the relevant reference band, and optionally applies the albumin correction when serum albumin is supplied.

Albumin is not on the BMP — order it separately.

The basic metabolic panel does not include albumin, yet albumin is the dominant unmeasured anion and drives the albumin-corrected gap. In any ICU or hypoalbuminemic patient, add a serum albumin to the order set so the corrected anion gap can be calculated. See the dedicated page on the role of albumin for the full Figge-equation derivation.

When the basic metabolic panel returns an abnormal anion gap, the next step is targeted testing guided by the differential — a venous or arterial blood gas for pH and PaCO₂, a serum lactate for Type A lactic acidosis, beta-hydroxybutyrate for DKA, and a serum osmolar gap when toxic-alcohol ingestion is suspected. Each of those add-on tests maps to a specific limb of the MUDPILES / GOLD MARK differential, converting a screening abnormality on the BMP into a confident diagnosis.