Interpreting an arterial blood gas (ABG) comes down to a consistent, step-by-step process: look at the pH first, then the carbon dioxide level, then the bicarbonate, and finally the oxygen. Each value narrows the picture, and by the time you’ve checked just the first two numbers, most of the interpretation is already done. The normal ranges you need to memorize are pH 7.35 to 7.45, PaCO2 35 to 45 mmHg, HCO3 22 to 26 mEq/L, and PaO2 75 to 100 mmHg.
Step 1: Start With the pH
The pH tells you the overall direction of the problem, and that direction doesn’t change no matter what the other numbers say. A pH below 7.35 means acidosis. A pH above 7.45 means alkalosis. If the pH falls within the normal range, note which end it leans toward: a pH of 7.36 is technically normal but sitting on the acidotic side, which matters when you look at the rest of the values.
Lock in that finding before moving on. Everything else you interpret will either explain the pH or explain the body’s attempt to fix it.
Step 2: Check the PaCO2
PaCO2 reflects how well the lungs are moving carbon dioxide out. Carbon dioxide is acidic, so a high PaCO2 (above 45) pushes the blood toward acidosis, and a low PaCO2 (below 35) pushes it toward alkalosis. The question to ask is simple: is the PaCO2 causing the pH problem, or is it trying to compensate for it?
If you identified an acidosis in step 1 and the PaCO2 is high, the respiratory system is the culprit. That’s a respiratory acidosis. But if the PaCO2 is low in the setting of acidosis, the lungs are blowing off extra carbon dioxide to compensate, which means the primary problem is metabolic. The same logic works in reverse for alkalosis. After just these two values, you can label most ABGs as primarily respiratory or primarily metabolic.
Step 3: Look at the Bicarbonate and Base Excess
Bicarbonate (HCO3) represents the metabolic side of the equation. If you already determined the problem is metabolic, the bicarbonate simply confirms it: low in acidosis, high in alkalosis. If you determined the problem is respiratory, the bicarbonate tells you something more useful: how long the problem has been going on.
In an acute respiratory problem, the kidneys haven’t had time to adjust, so bicarbonate stays near normal. In a chronic respiratory problem (one lasting days or longer), the kidneys retain or excrete bicarbonate to partially correct the pH. A significantly abnormal bicarbonate alongside a respiratory problem suggests the body has been compensating for a while.
Base excess (BE) appears on most ABG printouts and works alongside bicarbonate. A normal BE hovers around zero (typically negative 2 to positive 2). A strongly negative BE (called a base deficit) confirms metabolic acidosis. A strongly positive BE confirms metabolic alkalosis. It’s a quick confirmation of the metabolic picture you’ve already identified.
Step 4: Assess Oxygenation
PaO2 tells you how well oxygen is getting from the lungs into the blood. A normal PaO2 at sea level is 75 to 100 mmHg. Values below 75 suggest hypoxemia, and below 60 is generally considered significant enough to warrant supplemental oxygen.
When a patient is on supplemental oxygen, the raw PaO2 number loses context. A PaO2 of 80 on room air is normal, but a PaO2 of 80 on high-flow oxygen is alarming. The P/F ratio (PaO2 divided by the fraction of inspired oxygen) adjusts for this. On room air, where FiO2 is 0.21, a PaO2 of 100 gives a P/F ratio of about 476, which is healthy. A P/F ratio between 200 and 300 indicates mild impairment, 100 to 200 is moderate, and below 100 is severe. These thresholds are part of how acute respiratory distress syndrome (ARDS) is graded.
A quick shortcut to estimate whether oxygenation is appropriate: multiply the FiO2 by 500. The PaO2 should be roughly in that neighborhood. If it falls well short, there’s likely a problem with gas exchange in the lungs.
Identifying Compensation vs. Mixed Disorders
The body always tries to bring the pH back toward normal. In a metabolic acidosis, the lungs compensate by breathing faster to lower PaCO2. In a respiratory acidosis, the kidneys compensate by retaining bicarbonate. Compensation moves the pH toward normal but rarely corrects it completely. If the pH is perfectly normal but the PaCO2 and HCO3 are both abnormal, either compensation is very effective or you’re looking at two separate problems happening at once.
The key to spotting a mixed disorder is checking whether the compensation is appropriate. In metabolic acidosis, you can estimate the expected PaCO2 using Winter’s formula: multiply the bicarbonate by 1.5, then add 8. If the actual PaCO2 is higher than that predicted value, a respiratory acidosis is also present. If the actual PaCO2 is lower, there’s a simultaneous respiratory alkalosis. A simpler version that works well when bicarbonate is above 12: just add 15 to the bicarbonate value to get the expected PaCO2.
Another clue comes from the anion gap. In a high anion gap metabolic acidosis, the increase in the anion gap should roughly match the decrease in bicarbonate. If bicarbonate has dropped more than the anion gap has risen, a second metabolic acidosis (a non-anion-gap type) is also present. If bicarbonate hasn’t dropped as much as expected, there may be a coexisting metabolic alkalosis.
Calculating the Anion Gap
The anion gap helps narrow down the cause of a metabolic acidosis. The formula is: (sodium + potassium) minus (chloride + bicarbonate). A normal result is 4 to 12 mmol/L. Many clinicians drop potassium from the calculation for simplicity, which shifts the normal range slightly lower.
A high anion gap points to conditions where extra acid is being produced or not cleared: diabetic ketoacidosis, lactic acidosis from poor tissue perfusion, kidney failure, or certain toxins. A normal anion gap acidosis (sometimes called hyperchloremic) typically results from bicarbonate losses, as seen in severe diarrhea, or from certain kidney problems where acid isn’t excreted properly. Knowing which type you’re dealing with significantly changes what happens next clinically.
Common Causes of Each Disorder
Respiratory acidosis results from anything that reduces effective breathing: chronic lung disease like COPD or emphysema, sedating medications that suppress the drive to breathe, severe asthma attacks, or neuromuscular conditions that weaken the chest wall and diaphragm.
Respiratory alkalosis comes from excessive breathing. Anxiety and pain are common triggers. It also occurs when the body hyperventilates in response to low oxygen levels (at high altitude, for example) or as a compensatory response to an underlying metabolic acidosis.
Metabolic acidosis develops when acid builds up or bicarbonate is lost. Lactic acidosis from shock or sepsis, diabetic ketoacidosis, kidney failure, and prolonged diarrhea are the most frequent causes.
Metabolic alkalosis is often driven by volume loss: prolonged vomiting (which removes stomach acid), overuse of diuretics, or significant potassium depletion.
Sampling Errors That Skew Results
ABG values are only useful if the sample was handled correctly. Air bubbles trapped in the syringe are the most common source of error. Room air has a much higher oxygen concentration and lower CO2 concentration than venous or arterial blood, so an air bubble drives the PaO2 up (by roughly 10%) and the PaCO2 down (by roughly 5%). Even pH can shift slightly upward. If a result seems inconsistent with the clinical picture, a contaminated sample is worth considering.
Delays in processing also matter. Blood cells continue to consume oxygen and produce CO2 after the sample is drawn, so a syringe left sitting at room temperature for more than a few minutes will show a falsely low PaO2 and falsely high PaCO2. Samples should be analyzed promptly or kept on ice.
Venous Blood Gas as an Alternative
In many clinical situations, a venous blood gas (VBG) is drawn instead of an arterial sample because it’s less painful and easier to obtain. Venous pH runs about 0.03 lower than arterial pH, and venous PCO2 runs about 4 to 6 mmHg higher. For acid-base assessment, these predictable offsets make a VBG a reasonable screening tool. Where it falls short is oxygenation: venous PO2 reflects how much oxygen tissues have already extracted, not how well the lungs are delivering it, so it can’t replace an arterial sample for evaluating hypoxemia.