Can You Trust Your Watch's Blood Oxygen Reading at Altitude?
The moment a wrist-worn oxygen sensor is most useful, at real altitude, is also when the research shows it becomes least reliable.
This piece covers controlled-hypoxia validation studies of two consumer smartwatch SpO2 sensors against medical-grade reference devices, plus a real-world altitude study linking pulse oximetry to acute mountain sickness. It does not cover finger pulse oximeters or wrist SpO2 accuracy at sea level.
A watch's blood oxygen reading gets checked most anxiously exactly where oxygen actually runs short, at altitude, which makes its accuracy there matter more than anywhere else. Controlled testing of two consumer smartwatches under simulated high altitude found their SpO2 readings became increasingly unreliable as simulated altitude rose, with one study reporting an overall error rate above 50% and over 80% at the highest altitude tested, exactly the conditions where an accurate reading would matter most.
The one moment SpO2 actually matters, and the one moment it degrades
Most of the time, a wrist SpO2 reading is background information, a number that sits in the high 90s and gets ignored. At altitude, that same number becomes something people actually watch, because falling oxygen saturation is one of the real physiological signals behind altitude sickness.
That's precisely the scenario two separate controlled studies tested directly: taking a consumer smartwatch's SpO2 sensor into simulated high altitude and checking it against a real medical-grade reference.
2 studies
- A Garmin Forerunner 245 was tested against ECG and a medical oximeter across simulated altitudes from 3,000 to 5,500 meters. While heart rate readings stayed within an acceptable error range at nearly every altitude, the watch's SpO2 readings ran higher than the reference device overall, with an error rate above 50% across all tested altitudes and above 80% at altitudes over 4,800 meters, alongside poor reliability with the reference measurement.
- A Garmin fenix 5X Plus watch was compared against a medical-grade pulse oximeter across five simulated altitudes in an environmental chamber. Bland-Altman analysis found a 3.3% measurement bias for the watch's SpO2 readings at the highest simulated altitude tested, 12,000 feet.
Why the underlying signal still matters, even if the watch struggles with it
Neither study is arguing that oxygen saturation is meaningless at altitude, only that a wrist-worn sensor's read on it becomes shakier as conditions get more demanding. A separate field study on a real high-altitude ascent, 375 young men traveling by train from 400 meters to 4,086 meters over 43 hours, found participants who developed acute mountain sickness also showed a heart rate that ran significantly higher than those who didn't, alongside measured pulse oximetry, tying the underlying physiological signal to a real, self-experienced outcome rather than just a lab bias number.
That's the tension worth sitting with: the physiological signal a watch is trying to capture, falling oxygen saturation under real hypoxic stress, is genuinely connected to altitude sickness risk, even as the two controlled studies here show the watch's own reading of that signal degrades under the same conditions. Accuracy and usefulness pulling in opposite directions is a pattern this site has seen before in SpO2 research.
What neither controlled study tested
Both validation studies used environmental chambers to simulate reduced oxygen rather than a real mountain ascent, and both used small samples: 10 subjects in one, 13 women and 10 men in the other. Real altitude exposure adds factors like cold, exertion over hours or days, and acclimatization that a chamber test compresses into a single seated or lightly active session.
Neither study tested whether a wearer could still detect a meaningful downward trend in their own readings even if the absolute numbers ran inaccurate, which is a different, narrower question than overall measurement accuracy against a reference device.
Both smartwatch validation studies used normobaric environmental chambers with small samples (10 and 23 participants respectively) to simulate altitude over single test sessions. Neither establishes how these specific accuracy patterns hold up during a real, multi-day mountain ascent with cold, sustained exertion, and physiological acclimatization.
Common questions
Does a smartwatch's blood oxygen reading get less accurate at higher altitude?
In controlled testing, one study found a Garmin watch's SpO2 error rate rose from over 50% overall to over 80% at simulated altitudes above 4,800 meters, and a separate study found a 3.3% measurement bias at the highest simulated altitude it tested, 12,000 feet.
Is heart rate from a smartwatch also unreliable at altitude?
One study found heart rate readings stayed within an acceptable error range at nearly every tested altitude, even as the same watch's SpO2 readings became unreliable, suggesting the two sensors don't degrade in the same way under hypoxia.
Is oxygen saturation actually connected to altitude sickness risk?
In a field study of 375 men ascending to 4,086 meters, those who developed acute mountain sickness showed a significantly higher heart rate than those who did not, alongside measured pulse oximetry, connecting the underlying physiological signal to a real outcome.
Were these smartwatch studies done at real altitude or simulated?
Both validation studies used environmental chambers to simulate reduced oxygen levels rather than a real mountain ascent, which is a limitation the studies themselves don't extend to real, multi-day altitude exposure.
Sources
- Accuracy and Reliability of Commercial Wrist-Worn Pulse Oximeter During Normobaric Hypoxia Exposure Under Resting Conditions.
- Accuracy and Reliability of Pulse O2 Saturation Measured by a Wrist-worn Oximeter.
- Association Between Decreased Pulse Oximetry and Acute Mountain Sickness Upon Rapid Ascent to 4,086 m Among Young Chinese Men.