Plasma carries only about 1.5 to 3 percent of the oxygen in blood, and it does so purely by physical dissolution. Hemoglobin, packed into red blood cells, carries the other 97 to 98.5 percent by reversible chemical binding. So plasma does transport oxygen, but as a minor dissolved fraction rather than a delivery system. The real oxygen carrier is hemoglobin, and plasma is best understood as the solvent that lets red cells flow and the medium that moves carbon dioxide and heat.

Both statements in the question are partly true, which is why thin explainers keep getting the emphasis wrong. Plasma moves a small amount of oxygen without any chemistry at all. Hemoglobin moves nearly all of it through a molecular switch that responds to the tissue it is passing through. Understanding why that split exists explains everything from why carbon monoxide is so dangerous to why stored blood loses oxygen-carrying capacity faster than it loses volume.

The dissolution mechanism: why plasma can only carry so little

Oxygen is poorly soluble in water. At a normal body temperature of about 98.6°F, plasma holds roughly 0.003 milliliters of oxygen per 100 milliliters of blood for every 1 millimeter of mercury of oxygen pressure. That tiny coefficient is the whole story of plasma oxygen transport.

Arterial blood has an oxygen partial pressure near 100 mm Hg. Multiply 0.003 by 100 and you get about 0.3 milliliters of oxygen dissolved in each 100 milliliters of plasma. That is the entire plasma contribution, and it cannot be increased by breathing harder once hemoglobin is already saturated.

The solubility is fixed by physics, not by physiology. No enzyme, transporter, or pump in plasma can raise it. Water simply does not form bonds with oxygen the way heme iron does, so dissolved oxygen exists as isolated molecules bumping against plasma proteins and electrolytes. This is why hyperbaric oxygen therapy can push dissolved oxygen much higher, because it raises the partial pressure rather than the solubility.

That distinction matters because it separates two different jobs. Plasma oxygen is a pressure-driven pool with no storage capacity. Hemoglobin oxygen is a concentration-driven store with a huge capacity and a built-in release mechanism.

The hemoglobin mechanism: cooperative binding, not simple loading

Each red blood cell contains roughly 270 million hemoglobin molecules, and each molecule has four heme groups, so each cell can bind over a billion oxygen molecules. That capacity is what makes hemoglobin the dominant carrier, but capacity alone does not explain efficiency.

Hemoglobin binds oxygen cooperatively. When the first heme group picks up an oxygen molecule, the protein changes shape slightly and makes the next binding site more receptive. The fourth oxygen binds far more easily than the first. In the lungs, where oxygen pressure is high, this lets hemoglobin fill rapidly to near 100 percent saturation. In tissues, where oxygen pressure drops, the same cooperativity works in reverse, releasing oxygen steeply once pressure falls below a threshold.

The result is an S-shaped oxygen dissociation curve. The flat top keeps saturation high even if lung oxygen dips somewhat. The steep middle dumps oxygen quickly where tissues need it. A plain solvent cannot do this, which is the core reason plasma cannot substitute for hemoglobin.

A counterintuitive detail: hemoglobin does not just carry oxygen, it changes how much oxygen plasma can hold. By keeping the free oxygen pressure low in capillary blood, hemoglobin maintains the gradient that pulls oxygen out of plasma and into tissue. Without hemoglobin, dissolved oxygen would equilibrate with tissue but never reach the concentration needed for normal metabolism.

Quantified context: the actual numbers

The oxygen content of arterial blood is calculated by adding dissolved oxygen to hemoglobin-bound oxygen. A typical healthy adult has about 15 grams of hemoglobin per deciliter of blood, and each gram can bind about 1.34 milliliters of oxygen when fully saturated.

At 98 percent saturation, that works out to roughly 19.7 milliliters of oxygen per deciliter bound to hemoglobin. Add the 0.3 milliliters dissolved in plasma and total arterial oxygen content is about 20 milliliters per deciliter. Plasma's share is roughly 1.5 percent.

At a lower hemoglobin of 10 grams per deciliter, bound oxygen falls to about 13.1 milliliters per deciliter while dissolved oxygen stays near 0.3. Plasma's share rises to about 2.2 percent, but total delivery has dropped sharply. This is the key point: plasma's percentage share rises in anemia while its absolute contribution barely changes.

Component Oxygen carried per dL Share of arterial content Mechanism Can it be increased by breathing more?
Dissolved in plasma About 0.3 mL About 1.5% Physical dissolution No, once saturated, unless pressure rises
Bound to hemoglobin (15 g/dL) About 19.7 mL About 98.5% Reversible heme binding No, once saturated
Bound to hemoglobin (10 g/dL) About 13.1 mL About 97.8% Reversible heme binding No, once saturated
Dissolved under hyperbaric pressure Much higher, pressure-dependent Variable Physical dissolution Yes, by raising pressure

The takeaway from this table is that plasma oxygen and hemoglobin oxygen respond to completely different controls. One is governed by pressure and solubility, the other by protein chemistry and saturation.

Timeline: what happens across a breath and a capillary transit

The split between plasma and hemoglobin plays out over seconds. Tracking one red blood cell through a capillary shows how the two fractions interact.

Time point Location Plasma oxygen Hemoglobin oxygen What drives the change
0 seconds Pulmonary capillary Rising with alveolar pressure Loading as pressure rises High oxygen pressure in lungs
About 0.25 seconds Leaving lung Near equilibrium with alveoli Near 98% saturated Cooperativity speeds loading
About 1 second Arterial blood About 0.3 mL per dL About 19.7 mL per dL No change in transit
About 1 to 3 seconds Tissue capillary Falls as tissue consumes it Releases oxygen steeply Falling pressure and tissue demand
About 3 seconds Venous end Lower, partly depleted About 70 to 75% saturated Oxygen extracted by tissue
Resting whole-body Mixed venous blood Small residual About 70 to 75% saturated Balance of delivery and use

The important point is that plasma oxygen equilibrates almost instantly because dissolved gas diffuses freely. Hemoglobin oxygen lags slightly because binding and release take a fraction of a second, which is why the dissociation curve shape matters during fast capillary transit.

Edge cases: when the balance breaks

Several conditions expose how much the body depends on hemoglobin rather than plasma, and a few expose the reverse.

Carbon monoxide poisoning is the clearest example. Carbon monoxide binds hemoglobin with an affinity roughly 200 to 250 times that of oxygen, so it occupies binding sites and also shifts the curve so remaining oxygen releases less readily. Plasma oxygen is unaffected, which is why a normal oxygen pressure reading can mask severe poisoning. Pulse oximetry can also read falsely normal because it cannot distinguish carboxyhemoglobin from oxyhemoglobin in some devices. This is a red flag for clinicians.

Severe anemia shows the opposite pattern. With very low hemoglobin, total oxygen content falls, but dissolved oxygen in plasma stays roughly constant. The body compensates by increasing cardiac output and shifting the dissociation curve, but the margin is thin. A person can have a normal oxygen pressure and still deliver too little oxygen to tissues.

Red-flag warnings worth remembering: A normal oxygen saturation reading does not rule out carbon monoxide poisoning. Normal oxygen pressure does not guarantee adequate oxygen delivery in severe anemia. Hyperbaric oxygen raises dissolved plasma oxygen but is a pressure-based therapy, not a substitute for hemoglobin. Stored red blood cells lose some oxygen-releasing flexibility over time, so transfusion restores capacity but not always perfect function.

Misconceptions: what people get wrong

The most common misconception is that plasma transports no oxygen at all. It does, just in small amounts. Saying plasma carries zero oxygen is technically false and leads to muddled reasoning about hyperbaric therapy and dissolved gas physiology.

The second misconception is that plasma could replace hemoglobin if we simply dissolved more oxygen into it. The solubility limit makes this impractical at normal pressure, and the required pressures would be dangerous. This is why blood substitutes based on free hemoglobin or synthetic carriers have been pursued instead of simple oxygenated fluids.

A third misconception conflates oxygen content with oxygen pressure. Pressure tells you how much dissolved oxygen is present. Content tells you total oxygen in blood. They are related but not the same, and confusing them is a common source of clinical error.

The professional caveat: hemoglobin does not just carry oxygen, it also carries carbon dioxide and buffers acid. Plasma carries most carbon dioxide as bicarbonate, but hemoglobin handles a meaningful fraction and provides much of the blood's buffering capacity. Replacing red cells with a pure oxygen carrier would leave those jobs undone, which is one reason artificial blood has been so difficult to build.

Practical actions: how to reason about oxygen transport

You do not need to memorize dissociation curves to make good decisions, but a few concrete thresholds help.

First, interpret oxygen saturation with hemoglobin in mind. Saturation above 90 percent is generally acceptable in many settings, but saturation is a percentage of hemoglobin sites, not a measure of total oxygen content. A saturation of 95 percent with a hemoglobin of 7 grams per deciliter carries far less oxygen than the same saturation at 14 grams per deciliter.

Second, treat oxygen pressure and saturation as different numbers. A pressure near 100 mm Hg usually corresponds to high saturation in healthy lungs, but the relationship shifts with pH, temperature, and carbon dioxide. Do not assume one number predicts the other.

Third, when carbon monoxide is suspected, do not rely on standard pulse oximetry. Co-oximetry measures the actual hemoglobin species and is the appropriate test. This is a threshold action: any suspicion of exposure warrants the specific measurement, not a general saturation reading.

Fourth, in anemia, focus on total oxygen delivery rather than saturation alone. Delivery depends on cardiac output and hemoglobin concentration as much as on saturation. A normal saturation with low hemoglobin can still mean inadequate tissue oxygen.

Fifth, remember that hyperbaric oxygen works by raising dissolved plasma oxygen, not by increasing hemoglobin binding. This is why it can help in carbon monoxide poisoning and certain wound-healing situations, and why it is not a routine substitute for red blood cells.

Limitations: what this explanation cannot settle

The numbers above are typical adult values and vary with age, sex, altitude, and health status. Hemoglobin concentration ranges are broad, and oxygen-binding coefficients differ slightly between measurement methods. The 1.34 milliliters per gram figure is a standard estimate, not a universal constant.

The exact share of oxygen carried by plasma depends on the oxygen pressure at the moment of measurement, so quoting a single percentage is a simplification. At altitude, dissolved oxygen falls with pressure even while saturation stays high, which changes the picture. The clean 1.5 percent figure describes sea-level arterial blood in a healthy adult.

Finally, this explainer stays at the level of mechanism and general physiology. It does not provide diagnosis or treatment guidance for any individual. Decisions about oxygen therapy, transfusion, or hyperbaric treatment belong with qualified clinicians who can measure the specific patient's hemoglobin, oxygenation, and clinical status. The core answer remains stable: plasma does transport oxygen, but only as a small dissolved fraction, while hemoglobin does the heavy lifting through cooperative binding that no fluid can replicate.