A pulse oximeter shines light through your fingertip and measures how much of it gets absorbed. Blood that is carrying oxygen absorbs light differently from blood that is not, so by comparing two colors of light, the device can estimate what share of your blood is carrying oxygen. It does all of this without a needle, in a few seconds. That estimate is the SpO2 number on the screen.
This guide explains how that works, what the number is and is not, and how accurate the FDA and the UK’s medicines regulator, the MHRA, say these devices are. It sticks to what those bodies publish. We are not doctors, and nothing here tells you what your own readings mean.
What pulse oximetry is
Pulse oximetry is simply the test a pulse oximeter does. MedlinePlus, the US National Library of Medicine’s health site, describes it as “a test that uses a small, clip-like device called a pulse oximeter” and says it “is used to check your blood oxygen level.” It can be done at a provider’s office, a clinic, a hospital or at home.
The FDA describes the device itself as one that “estimates the amount of oxygen in the blood.” Note the word estimates. The FDA uses it every time, and it matters, as you will see below.
How the light does the work
The clearest explanation comes from a presentation the FDA gave to its expert advisory panel on pulse oximeters in February 2024. Here is the mechanism, step by step.
- Oxygen changes how blood absorbs light. Hemoglobin is the part of your red blood cells that carries oxygen. The FDA explains that hemoglobin with oxygen and hemoglobin without it “have different absorption spectra.” In plain terms, they soak up different colors of light by different amounts.
- The oximeter shines two colors through you. Small lights in the clip send light through your finger, and a sensor on the other side picks up what comes through. The FDA says the sensor measures the absorption of “2 or more wavelengths of light (typically red 660 nm and infrared 940 nm).”
- It looks for your pulse. The light also passes through bone, muscle, other tissue and the blood in your veins, and the FDA shows all of that as a steady background. Blood in your arteries is different: it surges with every heartbeat. The FDA notes that the absorption ratios “change with pulse with arterial blood,” so the device can pick out the part of the signal that rises and falls with your heartbeat. That is the arterial blood it wants to measure.
- A formula turns light into a percentage. The device compares how the red and infrared signals change, then converts that ratio into an SpO2 number “using algorithm and lookup table developed with calibration data,” in the FDA’s words. That is one reason the FDA asks for clinical testing to confirm a medical oximeter’s accuracy.
The MHRA puts the whole thing in one line: “these devices work by shining a light into the skin and measuring how this is absorbed by the blood.” Most pulse oximeters also show your pulse rate, which the FDA calls “similar to heart rate” and labels PR.
Not every oximeter shines light straight through. The FDA says oximeters for medical purposes include “devices using reflectance, transmittance, and fiber optic technologies,” and can be used at many sites, including the finger, ear, foot, hand, forehead, back and nose. Depending on the design, the sensor measures light that has passed through the skin or light that has been reflected and scattered back.
What SpO2 is, and is not
SpO2 is an estimate. The FDA describes it as “an estimate of how much oxygen the hemoglobin contains compared to how much it could contain, expressed as a percentage.” So 100% would mean all of the hemoglobin it can measure is carrying oxygen.
The real measurement needs a blood sample. The number SpO2 tries to match is called SaO2, the oxygen saturation of blood taken straight from an artery. The FDA says that the SaO2 measurement, “obtained via arterial puncture, is considered the gold standard.” MedlinePlus says the same from the patient’s side: “For a more accurate result, your provider may do a blood test called an arterial blood gas.”
It measures one thing only. The MHRA notes that pulse oximetry “does not measure carbon dioxide (CO2) levels.” The FDA adds that the method “does not account for presence of dysfunctional hemoglobin,” meaning hemoglobin that has been altered so it cannot carry oxygen properly.
How accurate pulse oximeters are
This is where the word estimate earns its place.
The international standard. The MHRA says pulse oximeters sold for clinical use are expected to comply with the standard BS EN ISO 80601-2-61:2019. Under it, the root-mean-square difference between the oximeter and the true value (often written Arms, a kind of average error) “must be below 4%” across the range of 70 to 100% SpO2. “Below 70% SpO2 the accuracy falls.”
What the FDA has asked for. In its 2024 presentation, the FDA said its 2013 guidance set an accuracy goal of 3% for transmittance sensors like fingertip clips and 3.5% for reflectance sensors. It also explained what that figure means in practice: a reading plus or minus the Arms value includes the true value about 68% of the time, and to cover about 95% of readings you double it. For a fingertip clip, that is plus or minus 6%.
What MedlinePlus tells patients. In simpler terms: “Pulse oximetry results may be 2% to 4% higher or lower than your actual blood oxygen level.”
None of this means oximeters are useless. The MHRA’s view is that “although pulse oximetry will not give a precise value of oxygen saturation, it is useful for following trends.” That is why every body we cite tells you to watch changes over time and how you feel, not one number. Our guide to using a pulse oximeter covers how to take a steady reading and when the NHS and FDA say to get help.
What can throw a reading off
Because the whole method depends on light reaching the sensor and blood pulsing under it, anything that blocks the light or weakens the pulse can affect the result. The FDA and MHRA list:
- Poor circulation or low perfusion, meaning reduced blood flow at the sensor site, and cold skin.
- Nail polish, false nails, dyes, henna and tattoos, which can block or change how light is absorbed. The MHRA suggests removing nail polish where possible, checking for false nails or using another site.
- Movement while the reading is taken.
- Bright light falling on the sensor, and a probe that fits badly. The MHRA’s rule: “Too tight causes constriction; too loose lets light in.”
- Skin pigmentation, skin thickness and current tobacco use.
- Severe anemia and conditions that alter hemoglobin, both on the FDA’s list.
Skin tone deserves its own explanation, because it follows directly from how the light works. The MHRA lists darker skin pigmentation among the factors clinicians should consider and says it “may cause an overestimate of SpO2 saturations,” meaning the oximeter shows a higher number than the true value. Our guide to pulse oximeter accuracy and skin tone covers what the FDA and the UK review found.
Medical device or gadget?
Every pulse oximeter uses light, but not every one has been checked. The FDA sorts them into two groups. Oximeters intended for medical purposes are reviewed through its 510(k) process, with “clinical testing used to confirm accuracy.” Others are sold for general wellness, sport or aviation, which the FDA says “are not intended for medical purposes” and are often sold directly to consumers online.
Since the device’s number comes from a formula built on calibration data, that testing is the only outside check that the formula gives the right answer. Our pulse oximeter picks include only models whose exact model number is named in an FDA clearance.
A pulse oximeter is clever, quick and painless, and it gives you an estimate, not a blood test. Use one made for medical purposes, take the reading carefully, and watch the trend along with how you feel. If you are worried about a reading or your symptoms, the FDA’s advice is to contact a health care provider.