How the sensor works
Most wrist-worn devices measure pulse with a technique called photoplethysmography, usually shortened to PPG. On the back of the device are small lights, commonly green and sometimes red or infrared, and a light detector beside them. The lights shine into the skin, and the detector measures how much light comes back.
Blood absorbs light differently from surrounding tissue. Each time the heart beats, a little more blood flows through the tiny vessels near the surface of the skin, and the amount of returned light changes slightly. Over time the detector records a gentle rhythm of rises and falls. Software finds the peaks in that rhythm and calculates how many occur per minute. That is your estimated heart rate.
Green light is common because it is absorbed strongly by blood and tends to give a clear signal near the skin’s surface, though designs vary. Some devices use several light colors to cope with different conditions.
From signal to number
The raw light signal is noisy. The device cleans it by removing slow drifts, filtering out frequencies that do not fit a heartbeat, and often comparing it to the motion signal from the accelerometer. If arm movement produces a rhythm that looks like a pulse, the software tries to subtract it. Many devices then average over a short window to give a steady number, which means displayed values can lag behind sudden changes.
Where the numbers are used matters. Resting pulse across the day, a workout summary and an alert for unusually high readings may each use different sampling rules. To save battery, many devices sample only now and then unless you start a workout, a trade-off related to sleep modes and duty cycling.
When readings are most reliable
Optical sensing tends to perform best when:
- you are at rest or moving gently and steadily,
- the device is snug against clean, dry skin, a little above the wrist bone,
- the skin is warm and blood flow near the surface is good,
- there is little ambient light leaking in around the sensor.
Under those conditions, many consumer devices track resting pulse reasonably well. Even then they are estimates and should not be assumed to match a clinical measurement.
When readings become less reliable
Several situations are well known for degrading wrist readings.
- Vigorous or irregular movement. Activities with rapid arm motion or gripping, such as weight training, rowing, boxing, or using handlebars, can disturb the signal. A device may lock on to the rhythm of your movement, such as your running cadence, instead of your pulse.
- Loose fit. If the band slides, light leaks in and the sensor loses contact. A snug fit is generally recommended during exercise.
- Cold. When skin is cold, the body reduces blood flow to the extremities and the signal weakens.
- Skin and tattoos. Dark ink, scars and some skin characteristics can absorb or scatter light. Research has discussed differences in performance across skin tones, which is one reason makers keep revising their designs, and a reminder of the broader topic of algorithmic bias.
- Hair, sweat and lotion. These can interfere with contact or light transmission.
- Rapid changes. During interval training, the displayed number can lag behind what your heart is doing.
- Irregular rhythms. Uneven heartbeats are harder for simple peak-finding methods to interpret.
- Sensor placement. Wearing the device too far toward the hand, or on the side with more movement, can change quality.
When the signal is poor, some devices show gaps, flat lines or obviously implausible values, while others quietly display a confident-looking number. This is why a sudden unexplained spike or dip should be treated skeptically.
Chest straps and other approaches
A chest strap measures the heart’s electrical activity rather than blood flow, and is often more accurate during fast, intense exercise. Some wearables, rings and earbuds also use optical sensing but from different places on the body; a finger or ear can provide a steadier signal than a moving wrist. Some watches also include electrical sensors, activated by touching a contact, for spot checks of rhythm. Those are intended for specific features and still do not replace medical assessment.
What the numbers are good for
Used sensibly, wrist pulse data can reveal patterns:
- how your resting rate drifts across weeks,
- how quickly your pulse rises during familiar effort,
- rough exercise intensity, when combined with a stable fit.
Resting pulse is influenced by sleep, caffeine, illness, stress, hydration and medication, so changes are not self-explanatory. The trends can be a prompt to pay attention, not a diagnosis. Alerts for unusually high or low pulse and notifications about irregular rhythm exist on some devices, but they have limits, can produce false alarms and can miss events. Anyone with health concerns should speak with a qualified professional instead of relying on a watch alone.
Practical tips
- Wear the device snugly, but not so tight that it is uncomfortable, and slightly higher than the wrist bone during exercise.
- Keep the sensor window and skin clean and dry.
- Warm up cold hands before relying on a reading.
- For hard intervals or strength work, consider whether a different sensor type suits you better.
- Check the app’s notes on how often it samples and what each mode does.
Pulse data is health-related personal information, so it is also worth looking at where it is stored and who can view it. That includes the apps connected to your account and any sharing options.