Can a Smartwatch Measure Blood Pressure?

The distinction that matters: only the first group measures blood pressure the way your monitor at home does. The second estimates it. The third never claims to.

The Huawei Watch D2 puts a genuine narrow inflatable cuff into the watch strap. Press start and it inflates against your wrist, exactly like a wrist monitor, then deflates. Because it is a real oscillometric measurement it can be validated like one, and independent testing has put it within a few mmHg of reference devices. It is sold across Europe, the UK, China and several Asian markets — but it does not hold US FDA clearance for blood pressure, so it is not a US option.

Omron's HeartGuide did the same thing years earlier and held FDA clearance, but it was expensive, bulky and limited as a watch, and Omron chose not to make a successor.

The trade-offs are real: these are chunky, the battery drains faster, inflating on your wrist in a meeting is conspicuous, and you still have to hold your wrist at heart level for the reading to mean anything. What you get in return is a number that behaves like a measurement rather than a guess.

This is the category most people picture. The watch shines light into your skin, reads how the reflected signal changes with each pulse, and an algorithm infers pressure from the waveform's shape and timing. Because that relationship is different for every person's arteries, the device must first be calibrated against a real cuff — and recalibrated as the relationship drifts.

• Samsung Galaxy Watch — calibrate with an upper-arm cuff at setup, then recalibrate roughly every four weeks. Availability is region-dependent and the feature has never been cleared for sale in the US.
• Hilo Band (Aktiia) — a wrist band rather than a watch, and the first cuffless monitor cleared by the FDA for over-the-counter use (July 2025). It calibrates over a four-day setup with a supplied cuff and recalibrates monthly, then takes readings through the day and night without inflating.

Independent reviews consistently find larger errors here than with cuffs — typically in the range of 5–10 mmHg for systolic, against ±3–5 mmHg for a validated arm cuff — and the error grows as you drift from your calibration point. These devices are best read as trend instruments: good at showing that your nights are worse than your days, or that something changed this month, and poor at telling you whether you are at 128 or 136 right now.

More detail in Cuffless Blood Pressure Monitoring: How It Works.

The Apple Watch does not measure blood pressure and Apple does not claim it does. Instead, on Series 9 and later and Ultra 2 and later running watchOS 26, it analyzes optical heart sensor data passively over 30-day windows, looking for signs that your blood vessels are responding to heartbeats in a way consistent with chronic hypertension. If it finds them, it tells you — and tells you to confirm with a real cuff over seven days.

It received FDA clearance in September 2025 and rolled out across more than 150 countries. It is not intended for people under 22, people already diagnosed with hypertension, or people who are pregnant.

This is a fundamentally different product: a screening tool aimed at the large number of people who have high blood pressure and do not know it. It will never give you a systolic number, and no notification is not the same as normal blood pressure. See Does the Apple Watch Measure Blood Pressure?

A large market of inexpensive "blood pressure smartwatches" displays a systolic and diastolic number that is not derived from any measurement of you in a meaningful sense. Typical tells:

• No calibration step of any kind — a device that has never seen your real blood pressure cannot estimate it
• Readings that are suspiciously stable, or that cluster near textbook values like 120/80
• No named regulatory clearance, no model number on any validation register
• Marketing that lists blood pressure alongside "blood glucose" and "blood oxygen" as equivalent features

Non-invasive glucose from a wrist optical sensor is not a solved problem, so a watch claiming it is a reliable signal about everything else on the box. Treat these numbers as decorative.

A wearable and a cuff are not competitors — they answer different questions, and the combination is more useful than either alone.

• Let the watch raise questions, and the cuff answer them. A run of high wearable readings, or a hypertension notification, is a prompt to run a proper 7-day cuff protocol.
• Keep calibration current. A cuffless estimate is only as good as its last calibration — that is the step people skip, and it is the step the accuracy depends on.
• Take cuff numbers to appointments. Treatment thresholds are defined on cuff measurements. Bring the log, note which device produced which readings.
• Watch the trend, not the digit. The wearable's strength is dozens of readings over weeks, including at night, which no cuff will ever give you.

"Medical-grade" is a marketing phrase, not a regulatory category. Nothing stops a listing using it, so it tells you nothing on its own. Two things do carry meaning: whether the device passed a clinical validation protocol, and whether a regulator cleared it to make a medical claim rather than a wellness one.

Three device classes fall out of that, none of which need a brand name to identify: a cuff worn as a watch, which can be validated like any other monitor; a calibrated cuffless estimator, which is a genuine technology with genuinely larger errors; and a passive screener such as Apple's hypertension notifications, which is cleared for what it does precisely because it never claims to produce a number.

You can settle this yourself in about five minutes, before buying anything:

• Search the independent registers by exact model name. validatebp.org covers devices sold in the US, and STRIDE BP maintains the international list used by the European and international hypertension societies. If the model is not listed, it has not passed an independent validation — the registers are opt-in, but any manufacturer serious enough to validate is serious enough to submit.
• Match the model number exactly. Manufacturers ship regional variants under near-identical names, and validation applies to the tested model, not to the family.
• Check the clearance in your own market. A device cleared in Europe may be unavailable or uncleared in the US and vice versa. The manufacturer's regulatory page should name the claim, not just say "certified".
• Read what the claim covers. Clearance for rhythm detection, or for a hypertension notification, is not clearance to measure pressure.
• Look for the calibration schedule. If a cuffless device asks for no calibration ever, it is not estimating your blood pressure from your body.

A blunt shortcut that is right more often than not: if a watch's blood pressure feature is on neither register and has no named clearance, treat its systolic and diastolic numbers as decorative and use a validated arm cuff for anything you would tell a doctor.

The other cuffless idea you will meet is the phone camera — hold a fingertip over the lens, or look at the front camera, and an app claims to return a blood pressure reading. The physics is the same optical signal a watch uses, with the same fundamental problem and none of the advantages: no contact pressure control, no calibration against your own cuff, and no clearance. Apps of this kind have repeatedly been removed from app stores for exactly that reason. See can a phone camera measure blood pressure? for the detail.

There is one genuinely useful thing a camera does here, and it is worth being precise about it: reading the numbers off a real monitor's display. That is optical character recognition of a measurement your validated cuff already made — not a measurement in itself. Blood Pressure Journal works this way: you take the reading with your cuff, point the camera at the screen, and systolic, diastolic and pulse are logged without typing. The measurement stays where it belongs, and the transcription stops being the part you skip.