How this instrument works
The alveolar-arterial (A-a) gradient measures the difference between the oxygen level the alveolar gas equation predicts should be in the lungs and the oxygen level an arterial blood gas actually measures in the blood. The calculation assumes three standard constants: 760 mmHg atmospheric pressure at sea level, 47 mmHg water vapor pressure at body temperature, and a respiratory quotient of 0.8. A small gap is expected and normal, since blood never picks up oxygen with perfect efficiency; a wide gap points toward something interfering with that handoff. StatPearls' physiology reference on the A-a gradient walks through the same reasoning this instrument automates.
A widened gradient and a low PaO₂ do not always mean the same problem. Simple underventilation — breathing too shallowly or too slowly — lowers PaO₂ and raises PaCO₂ together, but it does not widen the A-a gradient itself, because the alveolar prediction drops right along with the measured value. A gradient that widens while PaCO₂ stays put instead points toward a gas-exchange problem: blood passing through lung units that are poorly ventilated, poorly perfused, or bypassing gas exchange entirely. That distinction is the entire reason clinicians reach for this figure rather than reading PaO₂ alone.
The normal range is not one fixed number. On room air, a younger adult typically sits around 5-15 mmHg, and the expected range widens with age — StatPearls cites a rough estimate of (age ÷ 4) + 4 mmHg as the upper bound of normal. Raising FiO₂ toward 100% widens the expected gradient dramatically on its own, so a figure that looks alarming on room air can be unremarkable on supplemental oxygen, and vice versa. Sources also differ slightly on the water-vapor constant — 47 mmHg at 37°C is the figure used here and the more widely corroborated value, though at least one reference lists roughly 45 mmHg; the difference moves the result by well under a single mmHg and does not change how the gradient should be read.
- Enter FiO₂ as a decimal fraction — 0.21 for room air, up to 1.0 for 100% supplemental oxygen.
- Enter PaCO₂ in mmHg, read directly from an arterial blood gas report.
- Enter the measured PaO₂ in mmHg, from the same arterial blood gas draw.
- Read the A-a gradient in mmHg, then weigh it against the age- and FiO₂-adjusted normal range rather than a single fixed cutoff.
Worked example — room air, a lower PaO₂, and pure oxygen
Room air with a normal blood gas: FiO₂ = 0.21, PaCO₂ = 40, measured PaO₂ = 90. The bracket comes first: 0.21 × (760 − 47) = 0.21 × 713 = 149.73, then subtract PaCO₂ ÷ 0.8 = 40 ÷ 0.8 = 50, leaving 149.73 − 50 = 99.73. Subtract the measured PaO₂ of 90 and the gradient is 99.73 − 90 = 9.73 mmHg — squarely inside the 5-15 mmHg range generally considered normal for a younger adult.
Keep FiO₂ and PaCO₂ identical but drop the measured PaO₂ to 60: the bracket is unchanged at 99.73, and 99.73 − 60 = 39.73 mmHg. That much widening, on the same room air with the same PaCO₂, points toward a gas-exchange or diffusion problem rather than plain hypoventilation — hypoventilation alone would have pulled PaO₂ and the bracket down together and left the gradient close to normal.
Now switch to 100% oxygen with PaCO₂ = 40 and a measured PaO₂ of 300: the bracket becomes 1.0 × 713 − 50 = 663, and 663 − 300 = 363 mmHg. A gradient in the hundreds looks alarming next to the room-air range, but it is expected at this FiO₂ — the normal range itself has to shift upward whenever supplemental oxygen changes the bracket this much.
Questions
What counts as a normal A-a gradient?
Roughly 5-15 mmHg on room air for a younger adult, widening with age — StatPearls gives a rough estimate of (age ÷ 4) + 4 mmHg for the expected upper limit. There is no single universal cutoff; the figure has to be read against age and, separately, against whatever FiO₂ the person was breathing when the blood gas was drawn.
Why does the gradient get so large on 100% oxygen?
Because the predicted alveolar oxygen term scales directly with FiO₂, while achievable arterial oxygen does not rise as fast — small amounts of shunted or poorly matched blood flow have an outsized effect once the alveoli are saturated with oxygen. That is expected behavior, not a sign of a bigger problem than the same physiology would cause on room air; the normal range simply needs to be shifted upward at high FiO₂ rather than compared to the room-air figure.
What does a widened gradient on room air actually suggest?
That oxygen is not crossing from the alveoli into the blood as efficiently as expected, from causes like ventilation-perfusion mismatch, a diffusion barrier, or a right-to-left shunt. It argues against simple hypoventilation, since slow or shallow breathing alone lowers both the predicted and measured oxygen levels together and leaves the gradient close to normal rather than widening it.
Why 47 mmHg for water vapor pressure instead of another number?
47 mmHg is the standard value for water vapor pressure in fully humidified air at 37°C body temperature, and it is the figure most physiology references, including StatPearls, use in the alveolar gas equation. A small minority of sources instead cite roughly 45 mmHg; the two-mmHg gap changes the resulting gradient by well under one mmHg and does not meaningfully change how a result should be interpreted.
Can this calculator replace a clinician's interpretation of a blood gas?
No — it reproduces one specific piece of arithmetic from an arterial blood gas, not a diagnosis. A widened or normal gradient means something only alongside the rest of the clinical picture: oxygen saturation, the reason the gas was drawn, imaging, and the person's overall condition, all of which a clinician weighs together in a way a single number cannot.
References
Read this first: This instrument computes a screening figure from population formulas — it is not a diagnosis, and it cannot see the whole picture a clinician can. Use it to inform a conversation, not to replace one.