Fitbit’s SpO2 readings are reasonably good for spotting trends over time but not accurate enough to meet medical-grade standards. Research shows that wearable SpO2 sensors, including Fitbit’s, breach the accuracy thresholds set by the FDA and international medical device standards across all skin tones. That doesn’t make the feature useless, but it does mean you shouldn’t treat a single reading as a diagnosis.
How Fitbit Measures Blood Oxygen
Medical pulse oximeters clip onto your fingertip and shine light through your finger from one side to the other. Fitbit uses a different approach called reflectance pulse oximetry. Instead of shining light through tissue, Fitbit’s sensor sits flat against your wrist and measures how red and infrared light bounces back off your blood. The ratio of reflected red to infrared light changes depending on how much oxygen your hemoglobin is carrying, and Fitbit’s algorithm converts that ratio into an SpO2 percentage.
This reflection method is inherently less precise than the transmission method used by fingertip devices. Your wrist has more tissue, bone, and tendons between the sensor and your blood vessels, which introduces more noise into the signal. Fingertip oximeters also benefit from thinner tissue and dense capillary beds that produce a stronger, cleaner pulse signal.
What the Research Says About Accuracy
A 2024 systematic review and meta-analysis published in the Journal of Medical Internet Research evaluated SpO2 and pulse rate accuracy across wearable devices. The key finding: SpO2 measurements from wearables were inaccurate enough to breach both FDA and ANSI/AAMI/IEC standards. These standards typically require readings to fall within 2 to 3 percentage points of a medical reference device. Wearables, as a category, couldn’t consistently meet that bar.
The review also found that accuracy varied considerably depending on the specific model, likely reflecting differences in sensor hardware, algorithm development, and when the device was manufactured. Newer models with updated algorithms may perform better than older ones, but no consumer wrist-worn device matched the reliability of a clinical fingertip oximeter.
For practical purposes, this means a Fitbit reading of 95% could realistically represent a true value anywhere from roughly 92% to 98%. That range matters because 94% and below is generally considered clinically low, while 96% and above is normal. A single wrist reading can’t reliably tell you which side of that line you’re on.
What Affects Reading Quality
Several factors can push Fitbit’s SpO2 readings further from your actual blood oxygen level.
- Movement: Motion artifacts are one of the biggest accuracy killers. Red wavelength light penetrates deeper into tissue, making the pulse signal more susceptible to disruption from arm movements. Activities with sharp, random arm motions (like racket sports) produce the worst interference. This is why Fitbit primarily measures SpO2 during sleep, when your wrist is relatively still.
- Fit and strap position: A loose band lets ambient light leak in and reduces sensor contact. A band that’s too tight compresses the skin and restricts blood flow. Sweat during exercise can also make the strap slip, further degrading signal quality.
- Low perfusion: If blood flow to your wrist is reduced (from cold temperatures, dehydration, or certain medical conditions), the reflected light signal weakens and readings become less reliable.
- Skin tone: This is a nuanced finding. The 2024 meta-analysis found no statistically significant bias in wearable pulse rate readings across light, medium, and dark skin pigmentation groups. However, researchers have flagged the potential for bias in optical wearable sensors more broadly, and wider limits of agreement were observed in darker skin tones for some measurements. The concern isn’t settled science, but it’s worth knowing that the technology wasn’t originally optimized for all skin tones.
Differences Between Fitbit Models
All current Fitbit models with SpO2 capability, including the Charge 6, Sense 2, Versa 4, and Inspire 3, use the same basic reflectance sensor approach. The core health sensors (heart rate, SpO2, skin temperature) are shared across the lineup. There’s no evidence that one current Fitbit model produces meaningfully more accurate SpO2 readings than another, though newer models benefit from more recent algorithm updates that may improve consistency over time.
Where Fitbit SpO2 Is Useful
The real value of Fitbit’s SpO2 tracking isn’t in any single reading. It’s in the nightly trend. Because Fitbit measures blood oxygen during sleep (when motion artifacts are minimal and conditions are relatively consistent), the pattern over weeks and months can reveal meaningful changes. A gradual downward trend or recurring dips below your personal baseline could signal something worth investigating, like sleep apnea or respiratory changes at altitude.
Think of it as a screening tool, not a diagnostic one. If your Fitbit consistently shows SpO2 values below 92% during sleep, or you notice a sustained drop from your usual range, that’s useful information to bring to a doctor. But a one-off reading of 91% on a Tuesday night isn’t necessarily cause for alarm. It could reflect poor sensor contact, a cold wrist, or just sensor noise.
If you ever need a precise SpO2 reading for a medical reason, a fingertip pulse oximeter (available at most pharmacies for under $30) will give you a far more reliable number than any wrist-worn device currently on the market.