Every wearable ad shows the sensor list and the screen resolution. Almost none show the trade-off between them. The reality is that wearable battery life and sensor accuracy are two ends of the same rope. Pull one and the other moves. Once you understand why, you can pick a device that fits how you actually live rather than one that fits the marketing photo.
This is a walk-through of what the sensors cost, what the screens cost, and why the honest choice for a sleep-tracking user is often not the flashiest option on the shelf.
The single biggest drain: photoplethysmography
Almost every modern wrist wearable measures heart rate through photoplethysmography, or PPG. On the underside of the device, a set of green LEDs shine light into the skin, and photodiodes measure how much is absorbed as blood volume changes with each heartbeat. From that oscillating signal the device derives heart rate, heart rate variability, and, with additional wavelengths, blood oxygen saturation.
PPG is the biggest single drain on your battery. Every measurement means firing an LED, powering the photodiode, and running a signal-processing algorithm. The frequency at which the device does this — the sampling rate — is the biggest single lever a manufacturer has over battery life.
There are broadly three modes.
- Continuous, high-frequency sampling. The LEDs are on effectively all the time, or on for a full second every few seconds. Best signal quality; catches transient patterns like a sudden HR spike or a brief arrhythmia. Massively battery-hungry.
- Periodic sampling. The device wakes the sensor every minute, five minutes, or ten minutes, records a short window, and sleeps again. Good for trend data. Missable events between windows.
- On-demand sampling. The sensor only fires when the user opens the app or when a specific event is triggered. Cheapest by a mile. Loses all overnight data unless combined with a periodic mode.
When a manufacturer advertises a battery number, look for the footnote: with default settings. Default usually means periodic. Turn on continuous heart rate, and battery life shortens by half or more.
Continuous SpO2 is the silent killer
Blood oxygen measurement adds a red or infrared LED to the PPG stack because two wavelengths are needed to compute saturation. Continuous SpO2 monitoring — the kind that watches for sleep apnea signatures all night — costs even more battery than continuous heart rate, because both LEDs fire and the algorithm is more computationally demanding.
Almost every wearable that offers continuous SpO2 defaults it to off or to a limited overnight window. There is a good reason for that. Turn it on all night, every night, and you will feel the battery hit within a week. Turn it on during the day too, and some devices will not last a day.
Screens: the underrated cost
Sensors get the attention, but the display quietly eats a large fraction of your battery. Two technologies dominate.
AMOLED
AMOLED screens light each pixel individually. They look extraordinary in dark mode, produce true blacks, and are the standard for premium smartwatches — Apple Watch, Galaxy Watch, most Garmin flagships. The cost is that every lit pixel draws current. Always-on display modes on AMOLED devices are the single largest daytime battery expense, often larger than the sensor stack.
MIP transflective
Memory-in-pixel transflective LCDs work by reflecting ambient light rather than generating it. Under sunlight they look better, not worse. Under indoor lighting they look duller — that is the trade. Garmin has built much of its multi-week-battery reputation on MIP screens; the Instinct line, the Fenix series in some configurations, and the older Vivoactive models all lean on this technology. Newer AMOLED Garmins sacrifice weeks of battery to gain screen quality.
If you have looked at a Garmin next to an Apple Watch in a store and thought the Garmin screen looked washed out, that is not a defect. That is why it lasts fifteen times longer.
Cheap sensors: skin temperature and on-demand ECG
Not every sensor is expensive to run.
Skin temperature is basically free. A tiny thermistor takes a reading every few minutes and draws almost no current. This is why almost every mid-range wearable added skin temperature in the last three years without any hit to battery.
ECG is also cheap, because it is on-demand. You touch the crown or a metal contact for thirty seconds; the device records a single-lead trace and processes it; then everything shuts off. A weekly ECG check draws essentially nothing from your battery.
This is why the FDA-cleared ECG features on Withings, Apple, Fitbit, and Samsung watches are so widely available: the sensor itself is inexpensive to include and inexpensive to run. It is the always-on heart rate, always-on SpO2, and always-on display that shape the battery number.
Real numbers on shelves right now
Approximate battery figures for popular current devices, with defaults on unless noted.
- Apple Watch Series 9 and Ultra 2. Roughly 18 hours (Series 9) to 36 hours (Ultra 2) with always-on display, always-on HR, and periodic SpO2. Charges in about an hour on the puck.
- Withings ScanWatch 2. Approximately 30 days per charge. Hybrid analog dial with a small e-ink subdial; heart rate sampled periodically rather than continuously; SpO2 measured on demand rather than continuously overnight.
- Garmin Vivoactive 5. Up to 11 days in smartwatch mode with the AMOLED screen. A big drop from earlier MIP-based Vivoactives, but still an order of magnitude longer than an Apple Watch, because Garmin still sample-averages rather than continuously monitors most metrics.
- Garmin Vivosmart 5. Up to 7 days as a slim band with a small OLED display and periodic sampling.
- Fitbit Sense and Sense 2. Roughly 6 days with default settings, dropping into the 2 to 3 day range with continuous SpO2 and always-on display enabled.
- Whoop 4.0. Approximately 5 days per charge, with a battery pack that slides onto the band so you never have to take the device off. A subscription-only product; the battery pack workflow is a genuine engineering choice about how to trade continuous data collection against never having to charge on your wrist.
- Oura Ring 4. Roughly 4 to 7 days depending on activity. Sampling is periodic rather than continuous, and the ring form factor limits both battery volume and thermal envelope.
The range from 18 hours to 30 days is not a spec-sheet accident. It reflects fundamentally different decisions about what to measure, how often, on what kind of screen.
What you actually lose to save battery
The manufacturer of a 30-day watch is not lying about the battery. They are making a set of trade-offs that affect the data. Three of them matter most.
SpO2 gaps at night
Long-battery watches usually sample SpO2 periodically overnight — every fifteen or thirty minutes — rather than continuously. If you have moderate obstructive sleep apnea, this will still catch the pattern of desaturation cycles clearly enough to send you to a proper sleep study. If you have mild or positional apnea, some events between samples may be missed. A continuous SpO2 device catches more of them, at the cost of battery.
Missed atypical heart rhythms
Devices that sample heart rate every minute or five minutes will miss short paroxysmal arrhythmias that resolve within the sampling window. This is more of a theoretical loss than a practical one for most users — clinically significant AFib episodes generally last longer than a minute — but it is the trade being made. Continuous PPG plus algorithmic detection catches more transient events; that is why the Apple Watch's irregular rhythm notifications are marginally more sensitive than the periodic-sampling alternative.
Resampled HRV
Heart rate variability is derived from the beat-to-beat intervals over a window of time. Devices that sample continuously can compute HRV over any chosen window during sleep. Devices that sample periodically compute HRV during specific sensor-on windows and interpolate the rest. The absolute HRV values differ slightly across methodologies — which is a large part of why HRV numbers between brands are not comparable — and the trend fidelity of periodic sampling is a few percentage points lower.
The 30-day watch beats the 1-day watch (for sleep)
Here is the argument nobody in the industry wants to make out loud. For sleep tracking specifically, a wearable you actually wear every night is worth more than a wearable with perfect sensors that spends every third night on a charger.
Consumer wearables agree with polysomnography 60 to 75 percent of the time on stage classification. The absolute numbers on any given night are noisy for every device on the market. What is valuable is the trend across weeks and months. Trend accuracy is a function of two things: sensor quality and coverage. A 5 percent sensor advantage is easily wiped out by a 25 percent coverage gap because you took the watch off to charge.
This is why the honest recommendation for a sleep-first user is often a Withings ScanWatch 2, a Garmin device with several days of runtime, or an under-mattress mat — not the highest-spec sensor in the category. The one you charge every night is the one that misses a night of sleep every time you forget.
How to choose based on how you actually use a wearable
Three simple rules.
- If you want overnight sleep and HRV data primarily, pick a device with a week or more of battery, so that charging never interrupts your data. The Withings, Garmin, or Fitbit approaches all work here; the always-on-AMOLED approach does not.
- If you want maximum-signal cardiac screening and workout tracking, and you are willing to charge every one or two days, pick continuous PPG plus AMOLED — Apple Watch, Galaxy Watch, or an AMOLED Garmin. You will get slightly higher-fidelity data at the cost of coverage.
- If you want continuous SpO2 monitoring for suspected apnea, know that continuous SpO2 shortens battery on every device, and consider a non-contact under-mattress option that plugs into the wall instead. The measurement quality of a mat is competitive and it never needs charging.
The honest bottom line
There is no free-lunch device. Every wearable is trading sensor duty cycle against battery life against display quality against price. The specs sheet is a set of choices, not a set of features.
For most people who want a sleep-tracking or general health wearable and not a smartphone on the wrist, the best answer is a device with a week or more of battery, periodic sensor sampling, and a screen you can actually read outdoors. That is a wearable you keep on. And the wearable you keep on is the only one that produces the trend data any of these devices are actually good at.