The Apple Watch ECG app records a single-lead electrocardiogram in 30 seconds, then classifies your heart rhythm into one of several categories you can read right on screen. You don’t need medical training to understand what the result means, but you do need to know what the watch can and cannot tell you. Here’s how to take a reading, what each result means, and how to get the clearest trace possible.
How the Apple Watch Records an ECG
The watch uses two electrical contact points to detect your heart’s electrical signals. A sensor on the back crystal touches your wrist, and when you place a finger from your opposite hand on the Digital Crown, you complete an electrical circuit across your chest. The tiny voltage differences your heart generates with each beat travel through that circuit, and the watch plots them as a waveform over 30 seconds.
This is available on Apple Watch Series 4 and later, plus all Apple Watch Ultra models. The app is FDA-cleared as a Class II medical device, meaning it went through formal review for over-the-counter use. The current version (ECG 2.0) expanded the range of heart rates it can classify, covering rhythms from below 50 bpm up to 150 bpm.
How to Take a Reading
Open the ECG app on your watch. You’ll see a prompt to hold your finger on the Digital Crown. Before you do, make sure your band is snug enough that the back of the watch sits firmly against your skin. Rest your arms on a table or in your lap, stay still, and try to relax. The recording takes 30 seconds, and then you’ll see your result.
A few things that seem minor can ruin the reading. Your wrist and hands need to be completely dry: water, sweat, or residual moisture after washing your hands can interfere with the electrical signal. If you’ve been swimming or showering, Apple recommends waiting an hour or more for the watch to fully dry. Also move away from electronics plugged into an outlet, since they can introduce electrical interference. And make sure the watch is on the wrist you selected during setup, since the app uses that setting to orient the reading correctly.
What Each Result Means
After the 30-second recording, the watch displays one of several classifications. Here’s what you’re looking at:
- Sinus Rhythm (50 to 99 bpm): Your heart is beating in a normal, regular pattern. This is the result most people will see most of the time.
- High Heart Rate, No AFib (100 to 150 bpm): Your heart rate is elevated but the rhythm is regular. This could reflect exercise, caffeine, stress, or anxiety. It’s not flagging an arrhythmia.
- Atrial Fibrillation (50 to 150 bpm): The watch detected an irregular rhythm consistent with AFib, a condition where the upper chambers of the heart quiver instead of contracting in a coordinated way. This raises stroke risk over time and is worth discussing with a doctor.
- Low Heart Rate (below 50 bpm): Your heart rate is slower than the range the algorithm can classify. This is common in athletes or during sleep and isn’t necessarily a problem.
- High Heart Rate (above 150 bpm): Your heart is beating too fast for the algorithm to reliably classify the rhythm. The watch isn’t saying something is wrong, just that it can’t make a determination.
- Inconclusive: The algorithm couldn’t confidently categorize your rhythm. This can happen if you move during the recording, if your arm position shifted, or if the signal quality was poor.
- Poor Recording: The watch couldn’t get a usable signal at all. Usually this means the band was too loose, your skin was wet, or there was too much movement.
Reading the Waveform Itself
Beyond the classification label, the app shows you the actual ECG waveform and saves a PDF you can share with your doctor. You’ll see a repeating pattern of peaks and valleys for each heartbeat. The tall, sharp spike in the middle is the QRS complex, which represents the main contraction of your heart’s lower chambers. A normal trace shows these spikes at regular intervals with consistent spacing between them.
If you have AFib, the spacing between those spikes will look uneven, sometimes closer together, sometimes farther apart, with no predictable pattern. The baseline between beats may also look more chaotic instead of smooth. You don’t need to diagnose this yourself. The watch’s algorithm does the pattern recognition, and the PDF gives your doctor the raw data to confirm or override that classification.
To export the PDF, open the Health app on your iPhone, tap Browse, then Heart, then Electrocardiograms. Select a recording and tap “Export a PDF for Your Doctor.” This is a genuinely useful document for a clinician because it contains the actual waveform data, not just the watch’s interpretation.
How Accurate Is It?
The accuracy depends on what you’re measuring. In Apple’s own validation study of 588 subjects, the algorithm showed over 98% sensitivity and over 99% specificity for distinguishing sinus rhythm from AFib. That means it very rarely misses AFib when it’s present and almost never labels a normal rhythm as AFib.
Independent testing published in Circulation, a journal of the American Heart Association, found more nuanced results. When researchers had a physician interpret the saved PDF waveform, sensitivity for detecting AFib reached 96% with no false positives. But the watch’s own real-time notification caught only 41% of confirmed AFib episodes. The difference matters: the watch is better at recording AFib than it is at alerting you to it in the moment. If you suspect something is off, taking a manual ECG recording through the app gives you a much more reliable result than relying on passive notifications alone.
What the Watch Cannot Detect
This is the most important thing to understand. The Apple Watch ECG is designed to detect one specific condition: atrial fibrillation. It cannot detect a heart attack. It cannot identify blood clots, strokes, or other structural heart problems. A single-lead ECG captures electrical activity from one angle across your chest. A hospital ECG uses 12 leads placed at multiple positions around your torso, giving doctors a three-dimensional picture of your heart’s electrical activity. Many serious cardiac events only show up in leads the watch doesn’t have.
There has been early research exploring whether smartwatches could eventually detect heart attack patterns by taking multiple readings from different body positions, but as Harvard Medical School professor Dr. Peter Libby noted, that work is still “more a proof-of-principle rather than something that’s clinically useful.” If you’re having chest pain, the watch is not the tool to reach for.
Tips for Getting a Clean Reading
If you keep getting inconclusive or poor recording results, there are a few fixes worth trying. The most common culprit is moisture. Even a thin layer of sweat between the sensor and your skin can disrupt the signal. Dry your wrist and the back of the watch thoroughly before recording. Second, tighten the band one notch beyond what you’d normally wear. The back crystal needs firm, consistent contact with your skin throughout the full 30 seconds.
Body position matters more than people expect. Resting your arms on a flat surface like a table keeps them still and reduces the muscle noise that comes from holding your arms up or in an awkward position. Stay quiet during the recording. Even talking can introduce enough movement to degrade signal quality. If you’re near a laptop charger, power strip, or other plugged-in electronics, move a few feet away to avoid electrical interference that can show up as a jittery baseline on the trace.