Atrial fibrillation shows up on an ECG as three things happening at once: the normal, upright P waves that represent an organized atrial contraction disappear, the baseline between QRS complexes becomes wavy or jittery with chaotic fibrillatory activity, and the spacing between heartbeats turns irregular in a way that has no repeating pattern. Clinical diagnosis of AF relies on detecting these abnormal intervals between beats on an ECG tracing.1PubMed Central. Accurate detection of atrial fibrillation events with R-R intervals from ECG signals But the actual tracing can look surprisingly different from patient to patient, depending on how fast the heart is going, how much electrical chaos is present in the atria, and whether the signal is being filtered through an accessory pathway or affected by medication.
The Three Classic Features
On a normal ECG, every heartbeat starts with a small rounded bump called the P wave. That bump represents the atria contracting in a coordinated squeeze that pushes blood into the ventricles. In atrial fibrillation, the atria are not squeezing in any organized way. Instead, hundreds of tiny, disorganized electrical wavelets fire across the atrial tissue simultaneously. Because no single coordinated impulse is driving the atria, the ECG loses its distinct P waves entirely. What replaces them is a constantly shifting, irregular undulation of the baseline, sometimes described as a “sawtooth-less wobble” that looks nothing like the neat repeated waves you see in atrial flutter.
The second hallmark is the irregularly irregular ventricular rhythm. In most other heart rhythms, even abnormal ones, there is some regularity to the spacing between QRS complexes. Atrial fibrillation is different. The gaps between beats vary from one to the next in an unpredictable way. If you lined up a ruler along the R waves (the tall spikes on the tracing), you would see that none of the intervals match. This randomness is what clinicians mean by “irregularly irregular,” and it is one of the most reliable clues that AF is present rather than some other arrhythmia.
The third feature is the fibrillatory baseline itself. Between the QRS complexes, instead of a smooth, flat line interrupted only by T waves, the tracing shows continuous low-amplitude oscillations. These fibrillatory waves, often abbreviated as “f waves,” can be tiny and barely visible or large and prominent, depending on the patient. The QRS complexes themselves usually look normal in width, because the ventricles are still being activated through the usual conduction pathways. When they are not normal, that raises a separate set of questions covered below.
Why the Ventricular Rate Is So Unpredictable
The atria fire chaotically during AF, sending hundreds of electrical impulses per minute toward the ventricles. If every one of those impulses got through, the ventricles would beat at a lethal rate. They do not because the AV node, a small cluster of specialized cells sitting between the atria and ventricles, acts as a gatekeeper. It conducts slowly and has a long refractory period, meaning it needs time to recover after letting each impulse through before it can conduct another one.2PubMed. Integrated rate-dependent and dual pathway AV nodal functions: principles and assessment framework During AF, the ventricular rate is essentially determined by how quickly those AV node cells can recover and let the next impulse pass.3PubMed. Rate-dependency of action potential duration and refractoriness in isolated myocytes from the rabbit AV node and atrium
Because the incoming atrial impulses arrive at random intervals, some catch the AV node just as it recovers and get through quickly, while others arrive too early and are blocked. The result is a ventricular rate that typically lands somewhere between 100 and 170 beats per minute when AF is untreated, but with each individual beat-to-beat interval being different. Rate-controlling medications work by extending the AV node’s recovery time, which slows the ventricular rate and spaces out the QRS complexes on the ECG. Even with medication, though, the rhythm stays irregular. The intervals just get longer on average.
Fine Versus Coarse AF
Not all fibrillatory baselines look the same, and the difference matters more than you might expect. Clinicians divide AF into “fine” and “coarse” subtypes based on the height of those f waves on the ECG. Coarse AF shows larger, more visible undulations between QRS complexes, while fine AF has tiny, barely perceptible oscillations that can make the baseline look almost flat. The cutoff used in research is typically an f wave amplitude of 0.1 millivolts in standard leads.4PubMed Central. Coarse fibrillatory waves in atrial fibrillation predict success of electrical cardioversion
The distinction is not just cosmetic. Coarse fibrillatory waves on the surface ECG usually reflect more organized electrical activity within the atria, sometimes with patches of tissue that are nearly in flutter while the rest fibrillates. Right and left atrial recordings during coarse AF often show periodic episodes of slowing and near-regularity mixed in with the chaos.5American Heart Journal. Biatrial electrograms during coarse atrial fibrillation and flutter-fibrillation Fine AF, by contrast, tends to reflect a more thoroughly disorganized atrial substrate, which often tracks with more advanced atrial remodeling.
Practically, the fine-versus-coarse distinction shows up in treatment decisions. When patients with persistent AF undergo electrical cardioversion, those with coarse f waves are substantially more likely to remain in normal rhythm afterward. One study found that about 72% of patients with coarse waves maintained normal rhythm at six weeks, compared with 42% of patients with fine waves.4PubMed Central. Coarse fibrillatory waves in atrial fibrillation predict success of electrical cardioversion The thinking is that coarser waves mean the atria still have enough organized tissue to hold a normal rhythm once the chaos is interrupted. Fine AF may signal that too much of the atrial tissue has remodeled to sustain a coordinated beat.
Age also plays a role. Patients with fine AF tend to be significantly older than those with coarse AF.6Cardiology. Fine vs. Coarse Atrial Fibrillation: Which One is More Risky? In populations with rheumatic heart valve disease, fine morphology is far more common, affecting about 72% of AF patients with mitral stenosis, and age was the only factor independently associated with having the fine subtype.7PubMed Central. Fine versus coarse atrial fibrillation in rheumatic mitral stenosis: The impact of aging and the clinical significance On the other end, coarse AF has been linked to a higher rate of stroke. In a study of over 800 patients, roughly 20% of those with coarse AF had a history of cerebrovascular events, compared with about 14% of those with fine AF, and coarse morphology remained an independent predictor of stroke after adjusting for age.6Cardiology. Fine vs. Coarse Atrial Fibrillation: Which One is More Risky?
Rapid Ventricular Response on the ECG
When the ventricular rate during AF climbs above roughly 100 beats per minute, it is described as AF with rapid ventricular response, or AF with RVR. On the ECG, this shows up as QRS complexes that are packed tightly together, leaving little room for the fibrillatory baseline to be visible between beats. The tracing can look frantic, and the irregular spacing becomes harder to appreciate at first glance because everything is so compressed.
Interestingly, having a fast ventricular rate during AF may carry different long-term implications than you would expect. Registry data from a large Korean cohort found that AF with RVR was independently associated with a 39% lower risk of AF recurrence during follow-up, compared with AF at slower rates.8PubMed Central. Impact of Atrial Fibrillation with Rapid Ventricular Response on Atrial Fibrillation Recurrence: From the CODE-AF Registry The relationship followed a reverse-J shaped curve, meaning that moderate-to-fast ventricular rates predicted less recurrence than very slow ones. The likely explanation is that a fast rate during AF often reflects a less-remodeled AV node and less chronic atrial disease, so the heart is more capable of returning to and staying in normal rhythm.
When the QRS Goes Wide
Normally, the QRS complexes in AF look narrow and unremarkable because the ventricles are activated through the standard conduction system. But occasionally, a beat or a run of beats will suddenly look wide and different from the others. When that happens, clinicians have to figure out whether the wide beat is coming from the ventricle itself (a potentially dangerous ventricular ectopic beat) or is simply a normal impulse that took a detour through the ventricular muscle because part of the conduction system was still recovering from the previous beat.
The second scenario is called the Ashman phenomenon, and it is one of the more common sources of confusion when reading AF tracings. It happens when a relatively long gap between two beats is followed by a short gap. The long pause gives one branch of the conduction system extra time to recover, which paradoxically extends its refractory period. When the next beat arrives quickly after the short gap, that branch is not ready yet, so the impulse detours through the other branch. The result is a wide, right-bundle-branch-block-shaped QRS that looks alarming but is electrically benign.9BMJ Case Reports CP. Ashman phenomenon: a physiological aberration 10Journal of Emergency Medicine. Ashman’s Phenomenon: A Source of Wide-Complex Tachycardia
Telling Ashman beats apart from true ventricular ectopic beats matters because the treatments are very different. Unfortunately, the classic teaching that a long-short cycle sequence reliably identifies Ashman phenomenon is not as dependable as it sounds. One validation study using intracardiac recordings found that the long-short pattern was not specific for aberrant conduction and could also accompany genuine ventricular ectopic beats.11PubMed. Electrocardiographic criteria for differentiating aberrancy and ventricular extrasystole in chronic atrial fibrillation: validation by intracardiac recordings In practice, clinicians look at additional clues: the morphology of the wide beat, whether it resembles a typical bundle-branch block pattern, and whether there is a compensatory pause afterward.
Pre-Excited AF and Wolff-Parkinson-White Syndrome
There is one scenario where wide QRS complexes during AF are genuinely dangerous, and it has nothing to do with the Ashman phenomenon. In patients with Wolff-Parkinson-White syndrome, an extra electrical pathway connects the atria directly to the ventricles, bypassing the protective AV node. On a resting ECG during normal rhythm, this accessory pathway produces a characteristic triad: a short PR interval, a slurred upstroke on the QRS called a delta wave, and a widened QRS complex.12PubMed Central. Pre-Excited Atrial Fibrillation in Wolff-Parkinson-White (WPW) Syndrome: A Case Report and a Review of the Literature
If a person with this extra pathway develops AF, the chaotic atrial impulses can bypass the AV node entirely and pour directly into the ventricles at rates that would normally be filtered out. The ECG shows an irregular wide-complex tachycardia with variable QRS morphology, meaning the shape of each beat keeps changing because the impulses take different proportions of the normal and accessory pathways from beat to beat.13International Medical Case Reports Journal. Pre-Excited Atrial Fibrillation in a Young Woman with Wolff–Parkinson–White Syndrome Managed with Catheter Ablation Abroad and Followed in a Low-Resource Setting: A Case Report from Somalia The ventricular rate can be extremely fast. This is a medical emergency because it can degenerate into ventricular fibrillation. Standard AV-node-blocking drugs like beta blockers or calcium channel blockers can make the situation worse by forcing even more impulses down the accessory pathway.
Artifacts and Mimics That Fool the Machine
One of the trickiest aspects of reading ECGs for AF is that several non-cardiac phenomena can make a tracing look like AF when the heart is actually beating normally. Muscle tremor is the most common culprit. When a patient’s hands or limbs are shaking, whether from cold, anxiety, Parkinson’s disease, or age-related tremor, the small electrical signals from the muscle activity get picked up by the ECG electrodes and superimposed on the baseline. The result can look remarkably like fibrillatory waves.14Journal of the Indian Academy of Geriatrics. Multifocal Atrial Tachycardia Transitioning to Atrial Fibrillation in an Octogenarian
There are clues to spotting these pseudo-AF patterns. True fibrillatory waves tend to vary subtly in shape, amplitude, and timing across the tracing. Tremor artifacts tend to be more regular, produce sharply contoured waves rather than smooth undulations, and are more prominent in the limb leads than in the precordial (chest) leads. One case report highlighted several red flags: sharp, upright wave morphology that looked the same across all leads, different “flutter wave” shapes within a single lead, and a return to a flat baseline after each peaked wave, all of which pointed to artifact rather than true arrhythmia.15PubMed Central. Electrocardiographic artifact potentially misleading to the wrong management These mimics matter because misdiagnosis can lead to unnecessary blood thinners or heart-rate medications.
What Smartwatches Can and Cannot Show You
Consumer smartwatches have become a front-line AF detection tool for many people, and their ECG tracings look quite different from a clinical 12-lead. Most watches record a single-lead ECG strip, usually resembling lead I from a standard setup. You see QRS complexes, and if the rhythm is irregular enough, the watch algorithm flags possible AF. But the tracing lacks the spatial resolution to show fibrillatory waves clearly, and it cannot assess atrial activity in the same detail as a full 12-lead recording.
That said, the detection accuracy is surprisingly good for a consumer device. A systematic review pooling data from over 17,000 patients found that smartwatches overall achieved about 95% sensitivity and 97% specificity for detecting AF.16PubMed Central. Accuracy of Smartwatches in the Detection of Atrial Fibrillation: A Systematic Review and Diagnostic Meta-Analysis Performance varied by brand, with some devices favoring higher sensitivity and others higher specificity. A separate meta-analysis found that when algorithms read the tracings, sensitivity was around 86%, but when a trained human read the same smartwatch ECGs manually, sensitivity climbed to 96%.17PubMed Central. Accuracy and interpretability of smartwatch electrocardiogram for early detection of atrial fibrillation: A systematic review and meta‐analysis The practical takeaway is that a smartwatch flagging AF is worth taking seriously, but the watch’s “inconclusive” readings also deserve a follow-up rather than dismissal.
One major limitation is that AF is often intermittent. A standard 24-hour Holter monitor might catch only a fraction of episodes. One study comparing a 24-hour Holter to a 14-day continuous ECG patch found that the patch detected AF or atrial flutter in 22% of patients, while the Holter caught it in only 3%.18PubMed Central. Comparison of Arrhythmia Detection by 24-Hour Holter and 14-Day Continuous Electrocardiography Patch Monitoring Longer monitoring increases the odds of catching a paroxysmal episode, and wearable devices that continuously track heart rhythm sit somewhere in the middle of this spectrum.
Predicting AF From a Normal-Looking ECG
Perhaps the most surprising development in recent years is the discovery that artificial intelligence can look at an ECG recorded during normal sinus rhythm and predict that AF is likely to occur in the near future, even though the tracing looks perfectly ordinary to a human reader. A deep learning model trained on ECGs from US Veterans predicted the presence of AF within 31 days of a sinus rhythm ECG with an area under the curve of 0.86 in the main dataset, and 0.93 at an external validation site.19JAMA Cardiology. Deep Learning of Electrocardiograms in Sinus Rhythm From US Veterans to Predict Atrial Fibrillation
Another research group found that the AI model was picking up on subtle changes in the roughly quarter-second window just before each QRS complex, corresponding to the atrial activation period, where atrial remodeling leaves signatures too faint for the human eye to detect.20Scientific Reports. A new deep learning algorithm of 12-lead electrocardiogram for identifying atrial fibrillation during sinus rhythm The AI is not diagnosing AF that is happening right now; it is recognizing that the atrial tissue has already remodeled enough that AF episodes are occurring or imminent. This could eventually allow screening of patients during routine ECGs, flagging those who need extended monitoring long before they ever feel a symptom.
How the ECG Changes After Ablation
Catheter ablation is one of the most common treatments for AF, and it leaves its own footprint on the ECG. The procedure creates scar tissue around the pulmonary veins to electrically isolate the triggers that spark AF. Once the atria heal, the P waves that return after successful ablation look different from the ones that were there before the procedure. Specifically, research has found that the P wave in lead V1 shrinks after ablation. Its negative component gets smaller in area, and its positive portion gets shorter in duration and lower in amplitude.21PubMed Central. Change in P wave morphology after convergent atrial fibrillation ablation These changes reflect the fact that the left atrium’s electrical activation pattern has been physically altered by the ablation scars.
Clinicians have also started using these post-ablation P wave measurements to predict whether AF will come back. Several P wave parameters measured before and after the procedure, including maximum P wave duration and what is called the P wave terminal force in V1, have been shown to independently predict recurrence. Patients whose P wave characteristics changed more dramatically after ablation had different recurrence profiles, and researchers have built prediction models incorporating these measurements to help estimate individual risk.22PubMed Central. A Nomogram utilizing ECG P-wave parameters to predict recurrence risk following catheter ablation in paroxysmal atrial fibrillation Intracardiac spectral measurements during the ablation procedure itself can also forecast outcomes, with certain frequency patterns in the right atrium predicting a higher chance of AF returning.23PubMed. Surface ECG and intracardiac spectral measures predict atrial fibrillation recurrence after catheter ablation The ECG, in other words, is not just a diagnostic snapshot. It carries information about the atrial tissue’s structural health that keeps revealing new clinical insights as measurement tools improve.