Google already ships blood pressure on your wrist. What it ships is a trend, not a number, assembled from the sensors that are already there.
A newly published patent application describes hardware that could turn it into a measurement. And it doesn't work from the wrist at all.
What the Application Actually Is
The filing is US 2026/0272323 A1, published on September 17, 2026 and titled Computing Device Having Miniaturized Impedance Plethysmography Sensor for Measuring Blood Flow. It was filed March 14, 2025, carries application number 19/080,266, and runs to 20 claims. The applicant is Google LLC, and the record shows it was docketed and is ready for examination.
The inventor list is where the Google connection gets interesting: Seamus David Thomson, Seobin Jung, Debanjan Mukherjee, Ravi Narasimhan, Shwetak Patel and Samuel Wei Sheng, with addresses in Mountain View, Redwood City, Palo Alto, Los Altos, Seattle and Saratoga. Several of those names will be familiar to anyone who follows Google's health sensing work, and Google Research has published a paper on exactly this technique, titled Fingertip Impedance Plethysmography: A New Window Into Peripheral Blood Flow.
How It Works
The sensor is impedance plethysmography, or IPG, and the whole thing fits under one fingertip.
It uses four miniaturized electrodes. Two of them inject a small alternating excitation current into the finger tissue. The other two sit nearby and measure the resulting voltage. Blood conducts electricity differently from the tissue around it, so as the heart pumps and the blood volume in the finger changes, the measured impedance changes with it.
Run that measurement over time and you get a pulsatile waveform, a physical trace of the pulse. From there, the processor derives blood pressure, or as the application carefully puts it, a proxy for blood pressure.
That qualifier is in the patent's own language, not added by us, and it matters. It's the difference between a research claim and a medical one.
Why Not Just Use PPG?
This is the most interesting part of the document, because the patent explains its own reasoning in the background section.
The established optical approach is photoplethysmography, PPG, which is what every consumer wearable uses for heart rate. Google's filing argues fingertip PPG has real problems for blood pressure specifically.
PPG signals represent blood flow at the superficial tissue level, meaning the capillaries near the surface of the skin. That makes the signal sensitive to vasoconstriction and dilation, and to applied pressure. If you try to read hemodynamic information straight off a fingertip PPG signal, those confounders attack the accuracy of the blood pressure estimate.
The alternative is the oscillometric method, which applies pressure to the finger and analyzes the pressure-modulated signal. The patent points out that this runs into a different wall: under applied pressure it becomes difficult to distinguish arterial collapse from capillary collapse, which undermines reliability.
IPG sidesteps both by measuring the electrical response of the tissue rather than the light passing through it. That's the argument, anyway. Whether it holds up outside a laboratory is exactly what a patent application cannot tell you.
Where Google Thinks It Goes
The claims are explicit about how broadly this could be deployed.
Claim 16 lists the device types as a smartphone, smartwatch, tablet computer, laptop, earbuds, a headset, headphones, or gaming controllers. Figure 6A shows the sensor built into a watch body, with the sensing area positioned where a finger could rest against it. Others show finger-worn devices.
Claim 10 adds that the IPG sensor can be paired with a PPG sensor in an optical module positioned next to it, which suggests Google is thinking about combining both signals rather than replacing one with the other.
There is a second, separate Google application published the month before this one. It describes a small finger cuff with an inflatable bladder, an array of pressure sensors mapping pressure around the digital artery, and optical sensors measuring the pulse at the same time. That device is considerably less discreet than a fingertip sensor, and it reads more like a clinical instrument than something you'd wear to the gym.
Why This Is a Story and Not a Patent Round-Up
Because Google is already selling the watered-down version.
Blood Pressure Trends is arriving on Pixel watches this fall as part of Health Guardian, alongside insulin resistance trends and sleep breathing quality. That system does pattern analysis over weeks using sensors that are already on the watch. It gives you a direction of travel. It does not give you a reading, and Google is explicit that it isn't a prescreener for hypertension and shouldn't be used to adjust medication.
This patent is the first hard sign of Google working on hardware that would produce an actual pulsatile waveform rather than inferring from optical trends. That's the gap between what's shipping and what these filings describe.
Google's interest in the area is broader than one application. It ran a Fitbit Hypertension Lab study with Pixel Watch 3 owners last November, and an earlier Fitbit patent published in April hinted at blood pressure measurement from the wrist. Apple has been chasing the same problem by a different route, with a liquid-filled sensor it patented for future Apple Watch models.
The Caveats That Matter
A published application is not a product, and neither of these filings names a Pixel Watch or a Fitbit device. There is no timeline here, and there is no evidence either design has been built.
The filing date is March 2025, so this is not a reaction to anything recent despite being published this month.
And "or a proxy for blood pressure" is doing real work in that claim language. Nobody has cracked clinically accurate cuffless blood pressure in a wearable, and the companies that have come closest have done it with calibration and uncomfortable amounts of user discipline. A patent shows intent and a method. It does not show that the method works.
Bottom Line
Google is working on blood pressure from at least three directions now: optical trends from existing wrist sensors that are shipping today, an IPG sensor that reads blood flow through the fingertip, and a finger cuff with an actual inflatable bladder.
The fingertip IPG design is the most interesting of the three, because it's small enough to disappear into a watch or a phone, and because Google's own filing makes a specific technical argument for why it beats the optical method everyone else uses.
It's a patent, so it's a look at what Google is exploring and nothing more. But it's the clearest signal yet that the company isn't satisfied with selling you a blood pressure trend forever.
Sources: US patent application 2026/0272323 A1, "Computing Device Having Miniaturized Impedance Plethysmography Sensor for Measuring Blood Flow", applicant Google LLC, application 19/080,266, filed March 14, 2025, published September 17, 2026, including the abstract, claims 1 through 20, the background discussion of fingertip PPG confounders and oscillometric arterial versus capillary collapse, and the Figure 6A smartwatch embodiment; the earlier Google application for a finger cuff with an inflatable bladder, multiple pressure sensors and optical sensors, published August 2026; Google Research, "Fingertip Impedance Plethysmography: A New Window Into Peripheral Blood Flow" by Thomson, Jung and colleagues; Google's Health Guardian announcement and Blood Pressure Trends documentation for Pixel Watch 3, 4 and 5, covering the trend methodology and its stated limits; and Google's Fitbit Hypertension Lab study with Pixel Watch 3 owners, November 2025.
