Pulse oximetry has moved from hospital clipboards to wrist-worn gadgets, and the number it produces, blood oxygen saturation or SpO2, now appears in sleep reports, altitude dashboards, and workout summaries. The underlying technology is genuinely clever, but the wrist is one of the hardest places on the body to measure oxygen accurately. Understanding how the sensor works, and where it fails, keeps you from over-reading a single number.
What SpO2 Actually Measures
SpO2 is the percentage of hemoglobin in arterial blood that is carrying oxygen. At sea level, healthy adults typically sit between 95 and 100 percent, while 90 to 94 percent is considered low and anything below 90 percent is concerning. The important nuance is that a wearable never measures that percentage directly. It estimates it from how light travels through tissue, then converts an optical ratio into a number using a calibration curve. A medical pulse oximeter does the same thing with the same physics, but from a different location on the body and with different engineering constraints.
How a Wrist Sensor Uses Red and Infrared Light
Every optical SpO2 sensor relies on two wavelengths: red light near 660 nanometers and infrared light near 940 nanometers. Oxygenated and deoxygenated hemoglobin absorb those wavelengths differently. Oxygen-rich blood absorbs more infrared and lets more red light pass through, while oxygen-poor blood does the opposite. This is the same photoplethysmography principle behind smartwatch heart rate sensors, extended to a second wavelength.
The Ratio-of-Ratios Principle
The sensor shines both wavelengths into the skin, measures how much light returns, and calculates a value known as the ratio of ratios. That number is compared against a calibration curve built from volunteer studies in which healthy participants breathed controlled mixtures of air to lower their oxygen levels while a reference instrument recorded the truth. Because the curve comes from that narrow population, accuracy can drift for people who fall outside it, including those with poor circulation, unusual hemoglobin levels, or very different skin pigmentation. Regulatory agencies have specifically flagged pigmentation-related bias as an area of ongoing concern.
Pulsatile Signal Versus Static Tissue
Only a tiny fraction of the returned light carries the signal the sensor needs. Blood pulses with each heartbeat, so absorption fluctuates slightly, and that alternating component is what the algorithm isolates from the steady absorption of skin, bone, and non-pulsing venous blood. The useful signal may be one to two percent of total received light, which explains why motion, poor contact, or low blood flow can collapse the measurement entirely.
Why Fit and Wrist Placement Matter
Fingertip oximeters work well partly because fingers have dense capillary beds and the clip holds the sensor directly across a thin digit. The wrist has lower perfusion, more tendon and bone, and a sensor that must shine light through a thicker, less predictable path. A watch worn loosely, sitting over a wrist bone, or sliding during sleep changes the optical geometry from one reading to the next. Cold hands constrict peripheral vessels and reduce signal strength. Dark tattoos, wrist hair, and certain skin tones can absorb or scatter light and weaken the signal further. Placement is not a minor detail; it is the single largest source of error in wrist-based readings.
FDA Clearance Versus Wellness Features
Marketing language blurs an important distinction. Some wearables carry FDA clearance for a specific claim, such as measuring oxygen saturation on demand or estimating overnight blood oxygen for adults at rest. Other devices market SpO2 as a general wellness feature with no clearance at all. Even cleared features come with narrow intended uses, and a clearance does not turn a watch into a diagnostic device. Regulators have warned consumers not to rely on smartwatch oxygen readings to diagnose or rule out medical conditions, and that guidance applies regardless of brand or price.
Altitude, Cold, and Circulatory Conditions
Altitude legitimately lowers SpO2, and a reading of 90 percent at 12,000 feet can be normal rather than alarming. During acclimatization, oxygen saturation often dips at night while breathing patterns change, a pattern that overlaps with the territory covered in our guide to sleep apnea detection in wearables. Cold extremities, Raynaud’s phenomenon, low blood pressure, and conditions that reduce peripheral circulation can push readings down without any true drop in arterial oxygen. The reverse is also possible: a warm, well-perfused wrist can read reassuringly normal while a real problem develops elsewhere. The watch measures the wrist, not the whole body.
Spot Checks Versus Continuous Estimates
On-demand spot checks and background estimates are different products of the same sensor. A spot check holds the sensor still for 15 to 30 seconds to average a cleaner signal. Nighttime monitoring samples periodically, often once every few minutes, and reports a nightly range, an average, or a lowest value. Those overnight numbers come from far less data and far more motion, so a single low point carries little weight. Trends across weeks are more informative than any one night, and clinicians generally treat them as a prompt to investigate rather than a finding. If overnight oxygen is your priority, it is worth comparing wearable blood oxygen monitors built for nighttime tracking.
How Accurate Are Wrist Readings, Really?
Fingertip pulse oximeters must meet a standard of roughly three percentage points root-mean-square error against reference measurements to earn FDA clearance. Wrist devices vary far more widely. Independent studies generally find that wrist SpO2 tracks fingertip readings reasonably well when the wearer is still and warm, with errors often in the three to five percentage point range, and deteriorates sharply with motion, cold, or poor fit. Some studies have found that wrist devices can miss genuine hypoxemia, and others show them flagging low values that a fingertip device does not confirm. The practical takeaway is that wrist SpO2 is best treated as a trend line, not a precise measurement. Reviews of fitness trackers with accurate SpO2 readings run into exactly this limitation.
When a Medical Pulse Oximeter Is the Right Tool
Reach for a dedicated fingertip oximeter when the number matters: shortness of breath, chest pain, a known lung condition, altitude sickness, or monitoring during an illness. Fingertip devices are inexpensive, cleared as medical devices, and validated for the spot-check use case. For the cleanest reading, warm the hand, remove nail polish, rest the arm, and leave the device in place until the value stabilizes. If a wearable reading and a fingertip reading disagree, trust the fingertip device. If either suggests sustained low oxygen, especially with symptoms, contact a clinician. Arterial blood gas testing remains the clinical gold standard, and no consumer wearable comes close to replacing it.
Frequently Asked Questions
What is a normal SpO2 reading on a smartwatch?
For a healthy adult at sea level, 95 to 100 percent is typical. Values of 90 to 94 percent deserve a confirmatory fingertip check, and anything below 90 percent warrants prompt medical attention when symptoms are present.
Why does my watch report 88 percent at night?
Single overnight lows are usually artifacts caused by motion, poor contact, or cold skin. Repeat low values across several nights, or symptoms such as gasping or morning headaches, justify a conversation with a doctor.
Is wrist SpO2 as accurate as a finger pulse oximeter?
No. Wrist readings can come close when you are still and warm, but they degrade quickly with motion, cold, tattoos, or a loose fit. Treat them as directional rather than clinical.
Can a smartwatch diagnose sleep apnea?
No. Wrist oxygen trends can raise suspicion of breathing disturbances, but diagnosis requires a clinical sleep study. Devices that advertise apnea detection do so as screening signals, not diagnoses.
Does skin tone or nail polish affect the reading?
Pigmentation can influence accuracy because calibration studies skewed toward lighter skin. Nail polish interferes with fingertip oximeters, while on the wrist it is a non-issue. Tattoos and hair over the sensor can weaken the signal.
Why does SpO2 fail during workouts?
Exercise brings motion, sweat, and shifting peripheral blood flow, which break the already fragile pulsatile signal. Sensors need stillness to isolate the small alternating component from steady tissue absorption.
Should I trust my watch at high altitude?
Use it as a trend rather than an absolute. Expect lower readings at altitude, monitor how they change over days, and confirm concerning values with a fingertip device. Symptoms matter more than the number.
Are Apple Watch and similar devices medical grade?
It depends on the feature and the regulatory clearance in your country. Some devices hold clearances for specific oxygen measurements, but none are cleared for diagnosing or ruling out disease, and general wellness features carry no such clearance.