Apple Watch SpO2: a practical guide to when the reading is meaningful

TL;DR

Apple Watch includes a blood oxygen sensor in Series 6 and later. It uses reflectance pulse oximetry - a different and technically harder method than the finger-clip devices clinicians use. Published independent studies comparing Apple Watch Series 6 to medical-grade oximeters found 95% limits of agreement between ±2.7% and ±5.9% SpO2 depending on the population tested. That is a wider range than clinical-grade devices, which matters for medical decisions but is consistent with wellness use. Apple is explicit: Blood Oxygen app measurements are not intended for medical use and are only designed for general fitness and wellness purposes.

---

What the sensor actually does

Apple Watch uses four clusters of LEDs on the back crystal - a mix of red and infrared - along with four photodetector clusters. During a measurement, the LEDs shine light into the skin on your wrist. Some of that light is absorbed by haemoglobin in your blood; the rest is reflected back toward the photodetectors. The ratio of how much red versus infrared light is absorbed indicates how much of your haemoglobin is carrying oxygen - expressed as a percentage called SpO2.

This approach is called reflectance pulse oximetry. It is physically different from the method used by clinical finger-clip devices, which is transmissive pulse oximetry: light travels straight through the fingertip from one side to the other. Transmissive produces a stronger, cleaner signal because the light path is short, direct, and well-defined.

Reflectance from the wrist is harder. The light has to travel into the skin, scatter through tissue and capillaries, and return - a messier optical path that produces a weaker, more variable signal. Every bit of motion, every change in skin temperature or blood flow, adds noise. This is why the wrist location is technically more challenging for SpO2 than the fingertip, and why Apple Watch faces accuracy trade-offs that a hospital finger oximeter does not.

---

What the published accuracy data shows

A 2023 systematic review by Windisch and colleagues (published in Cureus, PMC10039641) pooled data from five studies involving 973 participants, all using Apple Watch Series 6. Comparing Apple Watch to medical-grade reference oximeters, the studies reported the following 95% limits of agreement:

The reviewers found no strong systematic bias - Apple Watch did not consistently overestimate or underestimate in one direction across studies. But the width of those limits of agreement is clinically significant. A ±5.9% band around any given reading means a true SpO2 of 90% could appear as anything from 84% to 96% in the worst-performing conditions. For tracking wellness trends over time, that is acceptable. For making clinical decisions, it is not.

The Rafl study - which tested participants under hypoxic (low-oxygen) conditions - produced the widest limits of agreement, which is worth noting for anyone thinking about using Apple Watch for altitude or respiratory monitoring purposes (more on that below).

---

Skin tone and accuracy

Pulse oximetry accuracy is affected by skin pigmentation, and this applies to all pulse oximeters - not only wearables. The mechanism is straightforward: melanin in the skin absorbs some of the same light wavelengths used by the sensor, which can cause the device to estimate higher oxygen saturation than the actual value.

The US Food and Drug Administration has consistently flagged this issue. Following an advisory committee meeting in November 2022, the FDA issued a November 2023 discussion paper on how to evaluate skin pigmentation effects in pulse oximeter testing. In January 2025, the FDA released draft guidance recommending that premarket testing for pulse oximeters specifically account for skin pigmentation diversity. A meta-analysis of clinical data cited by the FDA found that pulse oximeters overestimate SpO2 by a pooled mean of 1.11% in people with high skin pigmentation overall, and by 1.52% in Black and African American patients specifically.

For Apple Watch specifically, the picture is incomplete. Of the five studies reviewed by Windisch et al. 2023, only one (Pipek et al.) specifically examined skin colour as a variable - and found insufficient data to draw conclusions. The other studies enrolled predominantly Caucasian and white participants. This means the published accuracy figures for Apple Watch SpO2 do not yet adequately represent people with darker skin tones, and the actual accuracy in those populations may be modestly worse than the headline numbers suggest.

---

What Apple Watch actually does with blood oxygen

The Blood Oxygen app on Apple Watch provides two kinds of readings:

On-demand measurements take 15 seconds. You hold your arm still on a flat surface, tap Measure, and the watch returns a reading. Factors that can prevent a successful result include motion, a heart rate above 150 bpm, a loose watch fit, tattoos covering the back crystal, and poor skin perfusion (cold or clammy skin).

Background measurements happen automatically when you are not moving - primarily during sleep. These are recorded in the Health app and appear in your overnight Blood Oxygen history. The Vitals app introduced in watchOS 11 incorporates blood oxygen as one of its five overnight monitored metrics, alerting you when multiple metrics deviate from your typical range simultaneously.

Apple is direct about the feature's intended scope. The Blood Oxygen support page states: "Measurements taken with the Blood Oxygen app are not intended for medical use and are only designed for general fitness and wellness purposes." The feature is also restricted to users aged 18 and older.

---

The US availability note

Some US users have noticed that blood oxygen functions slightly differently on their Apple Watch. US models with part numbers ending in LW/A, purchased on or after 18 January 2024, process the SpO2 analysis on iPhone rather than on the watch itself, with results stored in the Health app. The feature still works as long as a compatible iPhone is present; the difference is where the computation happens. This reflects a hardware adjustment made in response to domestic patent constraints at that time.

---

When to trust it and when not to

The most honest framing comes directly from the accuracy data: Apple Watch SpO2 is a wellness signal, useful for identifying trends in your overnight readings over weeks and months - particularly in combination with other overnight metrics - but not reliable enough for point-in-time clinical decisions.

Use the data when you are looking for patterns. If your overnight blood oxygen has been consistently lower than usual across multiple nights, alongside elevated resting heart rate and disrupted sleep, that convergence is worth noting - a free app like Sam can surface it from your Apple Health data - and potentially worth discussing with a healthcare professional. A single outlier reading on one night is much less meaningful, since the measurement error alone can span several percentage points.

Do not use it to rule out respiratory illness, assess altitude acclimatisation, manage chronic conditions such as COPD, or make decisions about seeking medical care. These are use cases that require a validated clinical device - a medical-grade pulse oximeter or an arterial blood gas test - and a clinician to interpret the results.

If you are concerned about blood oxygen levels due to symptoms, travel to high altitude, or a known medical condition, use an appropriate clinical tool. Pharmacy-grade finger pulse oximeters cost under €30 and have been validated to clinical accuracy standards. Apple Watch does not replace them.

---

Blood oxygen at altitude

Readings drop at altitude because oxygen genuinely becomes less available - this is physics, not sensor error. At sea level, atmospheric oxygen pressure is high enough that the lungs easily saturate haemoglobin to 98-99%. As altitude increases, the partial pressure of oxygen falls even though the percentage of oxygen in air (21%) stays constant, so the lungs receive less oxygen per breath.

Research on healthy individuals at altitude shows:

• At approximately 2,900 metres (roughly the altitude of many ski resorts): mean SpO2 of about 93-94%
• At approximately 5,000 metres (high-altitude trekking, e.g. Everest Base Camp): mean SpO2 of approximately 80%
• Below 90% is generally considered the clinical threshold for hypoxaemia; symptoms of Acute Mountain Sickness tend to increase as SpO2 falls further

If you are at moderate altitude (1,500-2,500 metres) and your Apple Watch shows 94-96%, this is a normal physiological response. The relevant question is not "is this below 99%?" but "is this unusually low for this altitude, and am I feeling symptoms?" - shortness of breath at rest, severe headache, confusion, or persistent nausea at altitude warrant medical attention regardless of what a wrist sensor shows. The reading can also be more variable than at sea level, because reduced peripheral perfusion from cold mountain environments compounds the accuracy challenges described above.

---

Blood oxygen on a commercial flight

Aircraft cabins are not pressurised to sea-level equivalent - doing so would require heavier fuselage construction. Commercial aircraft are typically pressurised to an equivalent altitude of approximately 1,800-2,400 metres (roughly 6,000-8,000 feet). In healthy adults, this cabin pressure typically reduces SpO2 by 1-3 percentage points. A reading of 95-97% during a long-haul flight is physiologically normal and does not indicate a problem.

For people with pre-existing cardiorespiratory conditions (severe COPD, pulmonary hypertension, heart failure), reduced cabin oxygen pressure can be clinically relevant - but Apple Watch is not the right tool for making that assessment. Consult a physician before travelling if you have concerns about flight oxygen requirements.

---

Blood oxygen with a cold or respiratory illness

Common cold (rhinovirus): Typical rhinovirus infections cause upper respiratory inflammation - congestion, sore throat, runny nose - but do not significantly impair gas exchange in the lungs in otherwise healthy adults. Blood oxygen levels typically remain normal. If you see low readings during a cold, movement artefact or poor peripheral perfusion from fever are more likely explanations than genuine hypoxaemia.

Influenza and COVID-19: These illnesses can in some cases cause lower respiratory involvement that reduces gas exchange. Research during the COVID-19 pandemic identified SpO2 measurement as one of several physiological signals that shifted in some people with infection. A single low reading from a wrist-based device during illness should not be treated as a definitive sign of oxygen impairment - persistent breathlessness at rest is a more important signal than any number on your watch, and is worth discussing with a healthcare professional.

---

Where Sam Health fits in

Sam surfaces your Apple Watch overnight blood oxygen alongside your other overnight metrics - HRV, resting heart rate, wrist temperature, sleep duration, and respiratory rate - so a low SpO2 night appears in the context of everything else happening physiologically at the same time. A single dip in isolation rarely means anything. A pattern of low SpO2 alongside elevated heart rate and low HRV across multiple nights tells a more coherent story - including when you're travelling through altitude, spending time in the cold, or recovering from a respiratory illness. For a complete overview of the wearable metrics Sam works with, see the wearable biomarkers that actually matter, or explore the full picture of what Apple Watch sensors collect in our complete sensor breakdown for 2026.

Try Sam Health

Sources

• Use the Blood Oxygen app on Apple Watch - Apple Support, accessed 16 May 2026.
• Pulse Oximeters - US Food and Drug Administration, accuracy, limitations, and consumer guidance. Accessed 16 May 2026.
• February 2, 2024 Anesthesiology and Respiratory Therapy Devices Panel: Skin Pigmentation and Pulse Oximeter Accuracy - FDA Medical Devices Advisory Committee materials. Accessed 16 May 2026.
• Accuracy of Apple Smartwatch Sensors in Measuring Vital Parameters: A Scoping Review - Windisch C et al., Cureus, 2023; 15(3): e35843. PMC10039641. Accessed 16 May 2026.
• Roach RC, et al. High-altitude oxygenation. In: StatPearls. NCBI Bookshelf. NBK539701
• Coppel J, et al. SpO2 and heart rate during a real hike at altitude. Front Physiol. 2017;8:81. PMC 5303738