Pulse oximetry is one of the most commonly tested topics in nursing, respiratory therapy, and medical exams, and the questions usually hinge on a few key facts: how the device works, what it measures accurately, and where it fails. Understanding the core principles behind pulse oximetry makes it easy to evaluate any statement about it, whether on an exam or in clinical practice.
How Pulse Oximetry Works
A pulse oximeter uses two small LEDs that emit light at two specific wavelengths: red light at 660 nm and infrared light at 940 nm. These beams pass through a fingertip (or earlobe) and hit a detector on the other side. The device then calculates oxygen saturation based on how much of each wavelength is absorbed versus transmitted.
The key principle: oxygenated hemoglobin absorbs more infrared light and lets more red light pass through, while deoxygenated hemoglobin absorbs more red light and lets more infrared light pass through. The oximeter compares the ratio of red to infrared absorption and uses that ratio to estimate the percentage of hemoglobin carrying oxygen. This is displayed as SpO2. The underlying math is based on the Beer-Lambert law, which describes how light weakens as it passes through a substance that absorbs it.
Critically, pulse oximeters only analyze the pulsatile component of the signal, meaning the portion of light absorption that changes with each heartbeat. This is what isolates the arterial blood from the surrounding tissue, venous blood, and bone.
Normal Accuracy and What SpO2 Tells You
Medical-grade pulse oximeters are generally accurate to within plus or minus 2% when oxygen saturation is between roughly 76% and 100%. That means if your SpO2 reads 95%, your true arterial oxygen saturation is most likely between 93% and 97%. Below about 70% saturation, the devices become increasingly unreliable because they are calibrated using data from that higher range.
SpO2 correlates with PaO2 (the partial pressure of oxygen in arterial blood) along the oxyhemoglobin dissociation curve. This curve is S-shaped, and the flat upper portion acts as a buffer: PaO2 can drop from 100 mmHg all the way to about 60 mmHg before SpO2 falls below 90%. Once saturation dips below 90%, however, even small further drops in PaO2 cause large drops in SpO2. An SpO2 of 70% corresponds to a PaO2 of roughly 35 mmHg, which is dangerously low. This is why 90% is treated as a critical threshold in most clinical settings.
Carbon Monoxide Causes Falsely Normal Readings
One of the most commonly tested facts about pulse oximetry is that it cannot distinguish between oxyhemoglobin (hemoglobin carrying oxygen) and carboxyhemoglobin (hemoglobin bound to carbon monoxide). The two forms absorb light at very similar wavelengths, so the device reads carboxyhemoglobin as though it were oxyhemoglobin. In carbon monoxide poisoning, SpO2 may display a reassuringly normal number even while the blood’s actual oxygen-carrying capacity is severely reduced. This is called the “pulse oximeter gap,” the difference between the SpO2 reading and the true fractional oxygen saturation. Standard pulse oximeters will not alert you to this problem.
Methemoglobin Pushes Readings Toward 85%
Methemoglobinemia is another condition that fools pulse oximeters, but in a different way. Methemoglobin absorbs red and infrared light equally at the two wavelengths the device uses. As methemoglobin levels rise, the ratio of red to infrared absorption approaches 1:1, and the SpO2 reading trends toward 85% regardless of the patient’s actual oxygenation. If true oxygen levels are above 85%, the oximeter reads falsely low. If true levels are dangerously below 85%, the oximeter reads falsely high and stays stuck near that plateau. This makes pulse oximetry unreliable for monitoring methemoglobinemia in either direction.
Poor Perfusion Reduces Reliability
Because pulse oximeters depend on detecting a pulsatile signal, anything that reduces blood flow to the fingertip can degrade accuracy. Hypothermia, low blood pressure, vasoconstriction from cold or medications, and severe dehydration all reduce the perfusion index at the sensor site. When the perfusion index drops below 1%, errors become clinically significant.
A 2023 prospective study found that low perfusion combined with darker skin pigmentation dramatically increased the rate of missed hypoxemia diagnoses. Among patients with low perfusion, the oximeter displayed readings of 92% to 96% while true arterial saturation was below 88% in 1.1% of light-skinned subjects, 8.2% of medium-pigmented subjects, and 21.1% of dark-skinned subjects. In other words, the device was most likely to miss dangerously low oxygen levels in the patients who already had reduced blood flow to the sensor site and darker skin.
Skin Pigmentation Affects Accuracy
The FDA has acknowledged that pulse oximeters perform differently across skin tones. Current evidence shows the devices tend to overestimate oxygen saturation in people with darker skin pigmentation, sometimes by enough to mask true hypoxemia. In response, the FDA proposed updated recommendations for manufacturers, calling for more rigorous clinical testing across a range of skin tones during the device validation process. The agency partnered with academic institutions to conduct prospective studies in both adults and pediatric patients to better characterize these accuracy gaps.
A systematic review and meta-analysis found that while the overall root-mean-square error for individuals with dark skin pigmentation was about 1.88%, the 95% limits of agreement ranged from roughly negative 1.87% to positive 4.09%. That upper bound means some readings in darker-skinned patients could overestimate saturation by more than 4 percentage points, enough to hide clinically important hypoxemia.
Nail Polish and Other Interference
Nail polish can interfere with pulse oximetry because certain pigments absorb light at the same wavelengths the device uses. A systematic review found that black, blue, brown, and purple nail polish produced statistically significant drops in SpO2 readings, with black polish causing the largest effect. In one study, black nail polish prevented the oximeter from obtaining any reading at all in 88% of subjects, and brown polish blocked readings in 36%.
Green, orange, pink, red, white, and yellow nail polishes did not meaningfully affect SpO2 measurements. If you need a quick reading and removing nail polish is not practical, placing the probe sideways on the finger (side to side instead of top to bottom) can sometimes bypass the interference, though results vary.
Common Exam Statements: True vs. False
- “Pulse oximetry measures the partial pressure of oxygen in arterial blood.” False. It estimates hemoglobin oxygen saturation (SpO2), not PaO2. These are related through the dissociation curve but are not the same measurement.
- “Pulse oximeters use red and infrared light.” True. Red at 660 nm and infrared at 940 nm.
- “Pulse oximetry is reliable in carbon monoxide poisoning.” False. The device reads carboxyhemoglobin as oxyhemoglobin, producing falsely normal values.
- “SpO2 of 90% corresponds to a PaO2 of approximately 60 mmHg.” True. This is a widely tested correlation point on the oxyhemoglobin dissociation curve.
- “Pulse oximeters measure oxygen dissolved in plasma.” False. They measure oxygen bound to hemoglobin only.
- “Poor peripheral perfusion can affect pulse oximetry accuracy.” True. The device needs a detectable pulse to function properly.
- “Methemoglobinemia causes SpO2 to plateau around 85%.” True. Methemoglobin absorbs both wavelengths equally, driving the ratio toward a calculated saturation of 85%.
- “Pulse oximetry is equally accurate across all skin tones.” False. It tends to overestimate saturation in people with darker skin pigmentation.