Pulse Oximeters and SpO2 Estimates

Start here: symptoms matter more than the number

The FDA advises that if you use a pulse oximeter at home and are concerned about the reading, you should contact a health care provider, and that you should not rely only on the device; keep track of your symptoms and how you feel. [1] The FDA lists signs of low oxygen that include bluish color in the face, lips or nails, shortness of breath, a worsening cough, chest pain or tightness, and a racing pulse, and warns that some people with low oxygen may show none of them. [2] For severe breathing trouble or chest pain, call 911 in the US.

What a pulse oximeter measures

A pulse oximeter is usually clipped on a fingertip and uses light beams to estimate blood oxygen saturation (SpO2, shown as a percentage) and pulse rate. [2] The word to notice is "estimate". The device does not sample blood. Validation studies compare it with SaO2, oxygen saturation measured in a blood sample by CO-oximetry, which the review below treats as the gold standard. [3]

The FDA says SpO2 values are between 95% and 100% for most healthy people, can be lower in people with lung and heart problems, and are generally slightly lower at higher altitudes. [2] This is population information, not a personal target.

Why readings can be wrong

The FDA names several factors that can affect accuracy: poor circulation, skin pigmentation, skin thickness, skin temperature, current tobacco use and fingernail polish. [1] For home use it suggests a warm, relaxed hand held below heart level, no nail polish on the finger, sitting still and waiting for a steady number. [2]

Skin pigmentation: what the evidence and the FDA say

• FDA. The agency says current scientific evidence suggests some accuracy differences between lighter and darker skin pigmentation, and in January 2025 issued draft guidance (not final) on testing and labeling of medical-purpose oximeters across skin pigmentations. [2][1]
• Systematic review (32 studies, 6,505 participants). Pulse oximetry probably overestimates oxygen saturation in people with high skin pigmentation (pooled mean bias 1.11%, 95% CI 0.29 to 1.93%). The review noted that the extent of bias may be small in hospitals but is unknown in community settings. [3]
• Hospital records (30,039 paired readings, US Veterans Health Administration). A reading of 92% or higher hid blood oxygen below 88% in 15.6% of white patients and 19.6% of Black patients. [4] This is a retrospective record review, not a controlled test.
• Another review (23 oximeter studies). Accuracy values exceeded the 3% regulatory threshold for light, medium and dark pigmentation groups alike, so the authors concluded readings may be inaccurate across all groups. [5]

Label: established that accuracy is limited and differs by pigmentation in hospital data; insufficient evidence about size of the effect for home devices.

Watches and phones

A 2025 laboratory study tested a Samsung Galaxy S9/10 phone and a smartwatch (listed as "Galaxy 4") in 24 healthy volunteers at controlled arterial oxygen levels between 70% and 100%. Root mean square error was 2.6% for the phone and 1.8% for the watch, under the 3% standard in force at the time, and the authors said more testing in people with darker skin was needed. [6] It cannot be applied to other models, to resting daily use, or to overnight readings. Consumer devices that sell estimates as "wellness" features are not the same as a cleared medical oximeter. [2]

Accuracy is not benefit

None of the sources above show that tracking SpO2 at home improves health in healthy people. A reading that looks normal can sit alongside real low oxygen, as the hospital data show. [4] Overnight estimates cannot diagnose sleep apnea; see consumer sleep-apnea features and sleep apnea.

How to evaluate a claim

Ask whether the product is a cleared medical device or a wellness product, which exact model was tested, in whom, and against what reference. Also see fit, placement and skin factors and validation in different populations.