The top number on a pulse oximeter is your SpO2, which stands for peripheral oxygen saturation. It represents the percentage of hemoglobin molecules in your blood that are currently carrying oxygen, and in a healthy person at sea level it typically reads between 95% and 100%. That single number has become one of the most recognized vital signs outside of a hospital, yet what it really measures, how it can mislead you, and what counts as “normal” for different people are all worth understanding in more detail.
What the SpO2 Number Represents
Hemoglobin is the protein inside red blood cells that picks up oxygen in the lungs and delivers it throughout your body. Each hemoglobin molecule can bind up to four oxygen molecules. When the pulse oximeter displays “98%,” it means that about 98 out of every 100 hemoglobin molecules passing through the sensor are loaded with oxygen, and the remaining 2% are not. The relationship between oxygen levels in the blood and how much hemoglobin is actually saturated is not a straight line. It follows an S-shaped curve, meaning that small drops in blood oxygen tension at certain ranges can cause the saturation reading to fall quickly.
That S-shaped relationship, sometimes called the oxyhemoglobin dissociation curve, has practical consequences. When your SpO2 is in the mid-90s and above, oxygen is binding tightly to hemoglobin, and a modest dip in lung function might barely budge the number. But once the reading drops into the low 90s or high 80s, even a small further decline in oxygen tension can send the percentage tumbling. Several factors shift this curve, including body temperature, blood acidity, and carbon dioxide levels, all of which affect how readily hemoglobin grabs or releases oxygen.1PubMed. The oxygen dissociation curve: quantifying the shift This is one reason clinicians pay close attention when SpO2 starts drifting below 94%: the margin for error shrinks quickly from there.
How the Device Measures It
A pulse oximeter works by shining two wavelengths of light, one red and one infrared, through your tissue (usually a fingertip). Oxygenated hemoglobin absorbs more infrared light and lets more red light pass through, while deoxygenated hemoglobin does the opposite. A photodetector on the other side of your finger picks up the light that makes it through, and the device compares how much of each wavelength was absorbed.2PubMed Central. Pulse Oximetry with Two Infrared Wavelengths without Calibration in Extracted Arterial Blood
The clever trick is that the oximeter isolates the pulsatile component of the signal. Each heartbeat pushes a fresh wave of arterial blood into the fingertip, which briefly increases the tissue’s volume and changes the amount of light absorbed. By focusing only on that pulsing signal, the device filters out the light absorbed by bone, skin, venous blood, and other static tissue. This technique is called photoplethysmography, and it is also what allows the device to display your pulse rate, which is usually the second number on the screen.3PubMed Central. Current progress of photoplethysmography and SPO2 for health monitoring
The idea behind pulse oximetry traces back to 1972, when a Japanese engineer named Takuo Aoyagi was actually trying to measure cardiac output using an ear oximeter. He noticed that pulse-related noise in the red and infrared signals, which he originally wanted to eliminate, could itself be used to calculate oxygen saturation without needing to calibrate the device on each patient. That insight eventually became the basis for the commercial pulse oximeters that reached hospitals worldwide in the mid-1980s.4PubMed. Takuo Aoyagi: discovery of pulse oximetry
What Counts as a Normal Reading
For most healthy adults at sea level, an SpO2 of 95% to 100% is considered normal. Readings of 94% or below generally warrant attention, and anything under 90% is considered clinically significant hypoxemia, the point at which tissues may not be getting enough oxygen to function well. These thresholds are not arbitrary: they correspond to the steep portion of the oxyhemoglobin dissociation curve, where small drops in oxygen can cascade quickly.
During the COVID-19 pandemic, many people bought finger pulse oximeters for the first time and learned to watch for SpO2 readings below 94% as a reason to seek medical care. That threshold was reasonable guidance for otherwise healthy adults, but it does not apply universally. Some chronic lung conditions come with a permanently lower baseline, and treating those patients as though they need the same saturation targets as everyone else can actually cause harm.
When a Lower Number Is Actually the Target
People with chronic obstructive pulmonary disease, or COPD, are the most important exception to the “keep SpO2 above 94%” rule. In COPD, the lungs have difficulty expelling carbon dioxide, and giving too much supplemental oxygen can paradoxically suppress the body’s drive to breathe, causing dangerous carbon dioxide buildup. European and British guidelines recommend targeting an oxygen saturation of 88% to 92% in hospitalized patients with a COPD flare-up. Research has found that death rates were lowest in the 88% to 92% range, and even modest elevations above that, in the 93% to 96% range, were associated with a higher risk of death.5PubMed. Oxygen therapy and inpatient mortality in COPD exacerbation
This is not a small distinction. If you or a family member has COPD and you see an SpO2 of 90% on a home oximeter, that might be right on target rather than a cause for alarm. A randomized trial is currently underway to study whether pre-hospital emergency crews should also titrate oxygen to that 88% to 92% window in suspected COPD flare-ups, rather than giving high-flow oxygen by default.6PubMed Central. Standard vs. targeted oxygen therapy prehospitally for chronic obstructive pulmonary disease (STOP-COPD): study protocol for a randomised controlled trial The point is that “higher is better” is not always true for oxygen saturation, and the right target depends on the underlying condition.
Why the Number Might Be Wrong
Pulse oximeters are remarkably useful for a device you can clip onto a finger, but they have real limitations. Several factors can push the reading up or down from where it should be, sometimes by enough to matter clinically.
Skin Pigmentation
The most significant and well-documented source of pulse oximeter error is skin color. Because the device relies on light passing through tissue, darker skin pigmentation changes how much light is absorbed by tissue rather than by hemoglobin. The overall pattern is that oximeters tend to overestimate oxygen saturation in people with darker skin, meaning the device may display a reassuringly normal number when the true saturation is actually lower.7PubMed Central. A review of the effect of skin pigmentation on pulse oximeter accuracy
A large UK study of pulse oximeters used in COVID home-monitoring programs found that SpO2 readings were, on average, 0.6 to 1.5 percentage points higher in patients with dark skin compared to patients with light skin for the same actual arterial oxygen level. That may sound small, but it translates into dramatically different false-negative rates. Across the five oximeter models tested, the chance of a device reading above 94% when the true arterial saturation was at or below 92% was two to seven times higher in patients with dark skin than in those with light skin.8BMJ. The impact of skin tone on performance of pulse oximeters used by NHS England COVID Oximetry @home scheme: measurement and diagnostic accuracy study In clinical terms, this means a person with dark skin might look fine on the oximeter while actually needing supplemental oxygen or closer monitoring.
Cold Hands and Poor Circulation
Because the device depends on detecting a pulse of arterial blood in the fingertip, anything that reduces blood flow to the fingers can degrade the signal. Cold temperatures are the most common culprit. When your fingers are cold, blood vessels constrict, the pulsatile signal shrinks, and the oximeter has less information to work with. A study analyzing the effect of digit temperature on pulse oximetry found that cold conditions significantly reduced the quality of the light signal, lowering both its strength and reliability.9Biomedical Signal Processing and Control. Analysing the effects of cold, normal, and warm digits on transmittance pulse oximetry In children, a skin temperature below 30°C at the sensor site was identified as a significant predictor of oximeter inaccuracy.10PubMed. Effect of peripheral perfusion on accuracy of pulse oximetry in children
If your hands are cold and the oximeter is giving you a low or erratic reading, warming your fingers for a few minutes before retesting is a simple fix. A systematic review of oximeter performance in patients with poor peripheral perfusion found that most modern oximeters (about three-quarters of those tested) still performed accurately, and that earlobe sensors tended to be more accurate than fingertip sensors in these situations.11PubMed. Accuracy of pulse oximeters in measuring oxygen saturation in patients with poor peripheral perfusion: a systematic review
Nail Polish and Gel Manicures
Nail polish sits right in the path of the oximeter’s light beam, so it is natural to wonder whether it throws off the reading. The answer is: usually not by much. A systematic review covering 20 studies and 10 nail-polish colors found that black, blue, brown, and purple polishes produced small but statistically detectable dips in SpO2 readings. However, the differences were judged to be clinically insignificant in nearly all cases.12PubMed Central. Impact of Fingernail Polish on Pulse Oximetry Measurements: A Systematic Review One exception flagged by the review: in a single study, black nail polish caused the oximeter to fail to register a reading entirely in a high proportion of attempts. Gel-based manicures showed a similar pattern of small, mostly insignificant changes, though one brand of oximeter showed wide imprecision with certain gel colors.13PubMed Central. The effects of gel-based manicure on pulse oximetry If you are worried, simply rotating the sensor to a bare nail or using a different finger is the easiest solution.
SpO2 at High Altitude
If you have ever hiked above about 2,500 meters (roughly 8,000 feet), you may have noticed your oximeter reading dip into the low 90s or even the high 80s. This is expected. At higher elevations, there is less oxygen in the air, and your hemoglobin simply cannot saturate as fully. At the summit of a popular fourteener, an SpO2 in the mid-80s is not unusual for an acclimatized climber, and it does not necessarily mean something is wrong.
There are two caveats at altitude. First, the oximeter becomes less accurate when saturation drops below about 80%, because the calibration curves built into most devices are weakest at those low levels.14PubMed. Pulse oximetry at high altitude Second, there is no single “normal” SpO2 for a given elevation. Individual variation in acclimatization, lung function, and fitness means that what is fine for one person at 4,000 meters might be worrisome for another. Clinicians working in high-altitude medicine recommend interpreting the reading as a range rather than a precise number, and watching for trends over time rather than reacting to any single snapshot.
The Other Numbers on the Screen
Most pulse oximeters display at least two numbers. The top (larger) number is SpO2. The bottom number is your pulse rate, measured in beats per minute. The device derives the pulse rate from the same pulsatile light signal it uses to calculate saturation, counting each surge of arterial blood as a heartbeat. A normal resting pulse rate for adults is roughly 60 to 100 beats per minute, though well-conditioned athletes often run lower.
Some newer or hospital-grade oximeters display a third value: the perfusion index, or PI. This is a measure of how strong the pulsatile blood flow is at the sensor site, expressed as a percentage. A higher perfusion index means a stronger pulse signal, which generally means a more reliable SpO2 reading. When the perfusion index is very low, the device is struggling to detect pulses, and the SpO2 number should be interpreted cautiously. Factors that can drive PI down include cold extremities, low blood pressure, and peripheral vascular disease. Clinicians sometimes check the perfusion index before trusting an SpO2 reading, especially in intensive care settings.15PubMed Central. Peripheral perfusion index of pulse oximetry in adult patients: a narrative review
How Close Is SpO2 to the “Real” Number?
The gold standard for measuring oxygen saturation is an arterial blood gas, or ABG, which requires drawing blood from an artery and analyzing it in a lab. The pulse oximeter’s reading (SpO2) is an estimate of that true arterial value (SaO2). In most conditions, the two agree closely. One ICU study found that the average gap between pulse oximetry and arterial blood gas measurements was only about 0.2 percentage points, though individual readings could differ by as much as 6 points in either direction.16PubMed Central. Facing SpO 2 and SaO 2 discrepancies in ICU patients: is the perfusion index helpful?
A separate study in ICU patients confirmed that pulse oximetry readings tend to run slightly below the arterial blood gas values on average, and the gap widens when oxygen levels are low. When SpO2 was below 80%, the disagreement between the two methods grew meaningfully.17PubMed Central. Study of Oxygen Saturation by Pulse Oximetry and Arterial Blood Gas in ICU Patients: A Descriptive Cross-sectional Study For the vast majority of home users with readings in the 90s, the pulse oximeter is reliable enough to detect meaningful changes. When precision matters, such as in critical illness, blood gas analysis remains the definitive test.
Consumer Oximeters and the FDA
The fingertip pulse oximeters sold at drugstores and online for $15 to $40 occupy a regulatory gray area. Many are labeled “not for medical use” or “for wellness purposes only,” which means they have not been reviewed or cleared by the FDA. These devices are widely available and heavily marketed, but without FDA clearance, there is no guarantee they have been tested against the accuracy standards that hospital-grade devices must meet.18The Annals of Family Medicine. Clinical Interpretation of Peripheral Pulse Oximeters Labeled “Not for Medical Use” That does not mean they are all useless. Some may perform nearly as well as medical-grade units, but without standardized testing, you cannot know for certain which ones do.
If you are using a consumer oximeter at home, a few practical habits improve reliability. Take the reading while sitting still, with your hand at heart level and your finger warm. Wait for the reading to stabilize for at least 10 to 15 seconds rather than glancing at the first number that flashes. Remove nail polish if you can, or try a different finger. And pay more attention to trends over time than any single reading: a number that is gradually dropping across hours is more informative than one reading that looks slightly low.
Oximetry in Newborns and Young Children
Pulse oximetry is now used as a routine newborn screening tool for critical congenital heart defects, typically within 24 to 48 hours of birth. The sensor is placed on the baby’s right hand (which receives blood pumped before it passes through the ductus arteriosus, a temporary fetal blood vessel) and on one foot (which receives blood after that point). A significant difference in saturation between the two sites can signal a heart problem that needs immediate attention.
One study examined whether the left hand should be considered pre-ductal or post-ductal and found that oxygen saturations on the right and left hands were essentially identical, averaging about 95.7%, while foot readings ran about 0.8 percentage points lower on average.19PubMed Central. Pulse oximetry in the newborn: Is the left hand pre- or post-ductal? That small gap matters in screening protocols where clinicians are looking for differences of 3 percentage points or more between upper and lower body.
In small children, finger-sized sensors sometimes do not fit well, so clinicians may wrap the sensor around the sole of the foot or the palm of the hand. This works, but accuracy suffers when oxygen levels are low. One study in children with cyanotic heart conditions found that sensors on the sole of the foot were significantly less accurate and precise than sensors on the finger or toe, particularly when the true arterial saturation was below 90%.20Pediatric Critical Care Medicine. Pulse oximeter accuracy and precision affected by sensor location in cyanotic children Finger and toe placements performed more consistently across the full saturation range. For parents using a home oximeter on a young child, a toe is generally a more reliable site than the sole of the foot.
Carbon Monoxide and Other Hemoglobin Imposters
Standard two-wavelength pulse oximeters have a blind spot that is worth knowing about: they cannot distinguish oxygenated hemoglobin from carboxyhemoglobin, the form created when carbon monoxide binds to hemoglobin. Carbon monoxide locks onto hemoglobin roughly 200 times more readily than oxygen does, and when it is attached, the standard oximeter reads it as if it were oxygen. A person with dangerous carbon monoxide poisoning can show an SpO2 of 98% or higher on a regular pulse oximeter while their tissues are starved of oxygen.
Specialized multi-wavelength devices called pulse CO-oximeters can detect carboxyhemoglobin, but their accuracy drops off when oxygen saturation itself is low. One study found that carboxyhemoglobin detection was reasonably accurate when arterial saturation was above 85%, but below that level the device consistently produced errors and failed to report values.21PubMed Central. Accuracy of Carboxyhemoglobin Detection by Pulse CO-Oximetry During Hypoxemia Methemoglobin, another abnormal hemoglobin variant caused by certain medications and chemical exposures, similarly confuses standard oximeters. In cases of suspected poisoning or unusual hemoglobin levels, a blood gas analysis with co-oximetry is the only way to get a trustworthy picture.