How to Measure ST Elevation on an ECG

Measuring ST elevation on an ECG comes down to comparing the height of the ST segment against a reference baseline, using a consistent anatomical landmark on the tracing called the J-point. The technique sounds straightforward, but small choices about where you anchor the baseline and exactly where you measure along the ST segment can shift your reading by a millimeter or more, and in a clinical setting, a millimeter can determine whether a catheterization lab gets activated. The measurement itself is quick once you understand the landmarks, but the interpretation around it involves a surprising number of judgment calls.

Establishing the Baseline

Every ST-segment measurement is a comparison between two points on the ECG tracing: the level of the ST segment and the level of the baseline. If you pick the wrong baseline, every measurement that follows will be off. The most widely used reference point is the TP segment, the flat stretch between the end of the T wave and the start of the next P wave. When the heart rate is slow enough to produce a visible TP segment, this is the most electrically neutral portion of the cardiac cycle and makes a reliable zero line.

At faster heart rates, the TP segment shrinks or disappears entirely because the next P wave begins before the T wave fully resolves. In those cases, you shift to using the PR segment, the short flat line just before the QRS complex begins. The PR segment is not as electrically quiet as the TP segment, but it is nearly always visible, which makes it the practical default in tachycardic patients and continuous monitoring. Whichever baseline you choose, the key is consistency: use the same reference across every lead on the same ECG.

Finding the J-Point

The J-point is the spot where the QRS complex ends and the ST segment begins. On the tracing, it looks like the moment the rapid, steep deflection of the QRS abruptly changes direction and transitions into the more gradual slope of the ST segment. In some leads this transition is crisp and obvious. In others, especially leads where the QRS has a deep S wave that curves upward into the ST segment, the J-point can be subtle. A small notch or slur at that junction sometimes marks it clearly, but not always.

The J-point matters because it is your anchor for the actual measurement. You measure the vertical distance from your baseline (TP or PR segment) to the level of the ST segment at or just after the J-point. That vertical distance, expressed in millimeters on standard ECG paper (where one small box equals 0.1 millivolts, or 1 mm), is your ST elevation value.

Exactly Where After the J-Point to Measure

This is where practice gets slightly more nuanced than textbooks often let on. Different guidelines and different clinical contexts recommend measuring at the J-point itself, 20 milliseconds after it, 60 milliseconds after it, or even 80 milliseconds after it. The choice matters more than you might expect.

Research comparing these measurement points has found that the best diagnostic accuracy for detecting a heart attack tends to occur at or very close to the J-point. One study found that measuring at the J-point gave a specificity of 92% for acute coronary syndrome, and that the overall best measurement points were at the J-point or 20 milliseconds after it, though the optimal spot varied somewhat with age, sex, and whether the ST segment was elevated or depressed.¹ A separate study using ischemia-monitoring data found that measuring at 10 milliseconds after the J-point yielded the highest overall diagnostic accuracy at about 94%, with sensitivity and specificity both above 93%.²

The difference between measuring right at the J-point versus 20 milliseconds after it is usually small, averaging about 0.1 mm, but in about a quarter of patients the discrepancy is large enough to shift them into a different clinical category for ST-segment resolution after treatment.³ In practice, most current guidelines recommend measuring at the J-point. When you see older references mentioning “60 ms after the J-point” or “80 ms after the J-point,” those come from an era when measurement further along the ST segment was thought to be more stable. The trend has moved toward measuring closer to the J-point for better ischemia detection.

How Much Elevation Counts

Not all ST elevation means a heart attack. A small amount of elevation is normal in many healthy people, especially in the precordial leads V1 through V3. Current clinical criteria, as laid out in international consensus definitions, call for different thresholds depending on the lead, the patient’s age, and sex.

In most leads, 1 mm (0.1 mV) of new ST elevation at the J-point in two or more contiguous leads is the standard threshold. Leads V2 and V3 get higher thresholds because baseline elevation is more common there. For men aged 40 and older, the cutoff in V2–V3 is typically 2 mm. For men under 40, it goes to 2.5 mm because younger men frequently have more baseline ST elevation. For women, the threshold in V2–V3 is 1.5 mm. These sex- and age-adjusted cutoffs exist because applying a flat 1-mm rule across all leads produces too many false positives in the anterior chest leads.

The “two contiguous leads” requirement is critical. A single lead showing elevation might represent artifact, anatomical variation, or electrode placement issues. Two adjacent leads pointing in the same anatomical direction, such as II and III for the inferior wall or V3 and V4 for the anterior wall, give much stronger evidence that the elevation reflects a real event in a specific region of the heart.

Reciprocal Changes and What They Tell You

When you find ST elevation in a group of leads, look at the leads that face the opposite wall of the heart. ST depression in those “mirror” leads, called reciprocal changes, is a useful confirmation signal. In patients without altered conduction (no bundle branch block or pacemaker rhythm), reciprocal ST depression strongly suggests that the ST elevation represents a genuine heart attack rather than a benign cause.⁴

Reciprocal changes also carry prognostic weight. In one study, patients with ST elevation and reciprocal depression had significantly lower heart function and more extensive coronary artery disease compared to those with ST elevation alone. In the group with anterior heart attacks, those with reciprocal depression had multivessel disease about 80% of the time versus roughly half in those without it.⁵ So reciprocal changes are not just a diagnostic aid but a red flag for more severe disease.

Early Repolarization and How to Tell It Apart

One of the most common reasons for ST elevation in a young, healthy person is benign early repolarization. This pattern produces genuine elevation at the J-point, often with a small notch or slur, and it is concentrated in the precordial leads. The trouble is that it can look alarmingly similar to an anterior heart attack on a first glance.

Several features help distinguish the two. Early repolarization tends to produce widespread elevation rather than elevation limited to one coronary territory. The ST segment usually curves upward (concave shape), the T waves are tall and proportional to the QRS, and the pattern stays stable on serial tracings.⁶ In a heart attack, the R waves in the anterior leads tend to be smaller, the ratio of ST elevation to R-wave height is higher, and the QT interval is often longer. Researchers have derived a formula using the R-wave height in lead V4, ST elevation 60 ms after the J-point in V3, and the corrected QT interval that can distinguish the two patterns with about 88% accuracy.⁷ The practical takeaway: if the R waves are tall and healthy-looking in the leads where you see elevation, early repolarization becomes more likely. If the R waves are shrinking while the ST segments are rising, worry more.

Measuring Through Left Ventricular Hypertrophy

An enlarged left ventricle produces higher QRS voltages and often comes with baseline ST changes that can mimic or mask a heart attack. Standard ST-elevation thresholds perform poorly in patients with left ventricular hypertrophy (LVH) because the baseline is already shifted. One approach that improves accuracy is to measure the ST elevation as a ratio of the total QRS amplitude (the height of the R wave plus the depth of the S wave). In patients with anterior ST elevation and LVH, using a threshold of ST elevation equal to at least 25% of the R-to-S wave magnitude significantly improved the ability to identify a true blocked artery without losing sensitivity.⁸ Without this proportional approach, patients with LVH were frequently misclassified in both directions: real heart attacks were missed and benign patterns were flagged.

Left Bundle Branch Block and the Modified Sgarbossa Criteria

Left bundle branch block (LBBB) distorts the entire QRS-ST-T complex, making traditional ST-elevation criteria essentially useless. In LBBB, the ST segment normally deviates in the opposite direction from the main QRS deflection, a pattern called “appropriate discordance.” Trying to apply standard millimeter thresholds to this abnormal baseline leads to both missed heart attacks and false alarms.

The original Sgarbossa criteria addressed this by identifying specific patterns that suggest a heart attack despite LBBB, but their sensitivity was limited. A modified version, sometimes called the Smith-modified Sgarbossa criteria, replaced one of the original rules with a proportional measurement: if the ST elevation in any lead is at least 25% of the depth of the preceding S wave (an ST-to-S ratio of −0.25 or beyond), the pattern is suspicious for a heart attack. This modified rule raised sensitivity to about 91% while keeping specificity around 90%, a substantial improvement over the original criteria.⁹ A separate validation study confirmed the improvement, finding that the modified criteria were significantly more sensitive than either version of the originals while maintaining very high specificity.¹⁰ Paramedics have used these modified criteria in the field to activate catheterization labs directly.¹¹

Diffuse ST Elevation From Pericarditis and Metabolic Causes

Not all ST elevation follows a coronary territory. Pericarditis, inflammation of the sac around the heart, produces widespread, concave ST elevation that appears in many leads simultaneously rather than mapping neatly to one arterial distribution. A hallmark clue is PR-segment depression in the limb leads and PR elevation in aVR, a pattern that does not occur in heart attacks.¹² The diffuse, concave, multi-lead distribution is the main differentiator when measuring ST elevation in this context.

Severe electrolyte disturbances can also produce ST elevation. Extremely high potassium levels sometimes create a pattern called “pseudoinfarction,” where the ECG shows ST elevation that looks just like a heart attack but resolves once the potassium is corrected. The term “dialyzable current of injury” reflects the fact that this ST elevation disappears with treatment, reinforcing the importance of checking basic labs before committing to an invasive strategy based solely on the ECG tracing.¹³

Technical Pitfalls That Create False Elevation

Before you interpret any ST measurement, make sure the ECG machine itself is not fooling you. Two common technical problems can manufacture ST elevation out of nothing.

The first is filter settings. ECG machines use high-pass filters to reduce baseline wander, but when these filters are set too aggressively, they distort the ST segment. A high-pass filter at 0.5 Hz, the typical telemetry default, can produce clinically significant false ST elevation, especially in the right precordial leads (V1 and V2) and in patients with wide QRS complexes.¹⁴ At a 1.0 Hz filter setting, the odds of a false-positive ST elevation of 1 mm or more increased nearly tenfold in V1 and fivefold in V2.¹⁵ The fix is straightforward: always use diagnostic-mode filter settings (0.05 Hz high-pass) when you are evaluating ST segments. Monitor-mode or telemetry-mode tracings are not reliable for this purpose.

The second pitfall is lead misplacement. Swapping the right and left arm electrodes, for instance, can flip the polarity of certain leads and create a pattern that mimics ST elevation in a coronary territory.¹⁶ If the ECG looks alarming but something about it feels off, checking electrode placement and repeating the tracing takes seconds and can prevent a false catheterization lab activation.

When the Standard 12 Leads Are Not Enough

The standard 12-lead ECG does not directly view the posterior wall of the heart or the right ventricle. If you see ST elevation in the inferior leads (II, III, aVF) and suspect an inferior heart attack, the right ventricle and the posterior wall are blind spots that the standard leads cannot assess reliably.¹⁷

The solution is to add leads. Lead V4R, placed on the right side of the chest in the mirror position of V4, reflects the electrical activity of the right ventricle. Leads V7, V8, and V9, wrapped around the left side of the back, image the posterior wall. Together with the standard 12, these form the 15-lead ECG, which is the most commonly used expanded configuration.¹⁸ Posterior and right ventricular infarctions are likely underdiagnosed precisely because these extra leads are not recorded routinely. Any time you measure ST elevation in the inferior leads, obtaining a right-sided and posterior ECG adds only a minute or two and can reveal an infarction that would otherwise be missed.

How Much to Trust the Machine’s Reading

Modern ECG machines print an automated interpretation at the top of the tracing, and in many emergency systems this interpretation is what triggers the initial alert. The machine’s accuracy is decent but far from perfect. In one study of confirmed STEMI patients, the automated interpretation was wrong about a quarter of the time, with a misinterpretation rate of roughly 27%.¹⁹ A comparison of three different automated algorithms found sensitivities ranging from 62% to 69%, meaning roughly a third of true heart attacks were missed by at least one algorithm.²⁰

The machines perform better on specificity, generally above 89%, so when the machine says “STEMI” it is usually right. The danger lies in false negatives: the machine reading “normal” or “nonspecific changes” when there is actually significant ST elevation. For prehospital settings, a meta-analysis of paramedic-transmitted ECGs found pooled sensitivity around 85% and specificity around 95% for computer-assisted STEMI identification.²¹ The practical lesson: always look at the tracing yourself rather than reading only the machine’s printout, and measure the ST segments manually if the clinical picture suggests ischemia.

Transient and Fleeting ST Elevation

ST elevation does not always persist long enough to be captured on a standard ECG. In coronary artery spasm (Prinzmetal angina), the ST segment rises abruptly during the spasm and resolves within minutes as the vessel relaxes. About half the time, the first visible change is a tall, peaked, symmetrical T wave, followed by progressive ST elevation that builds and then fades as the spasm passes.²² Because of this brief window, a normal resting ECG does not rule out spasm-related ST elevation. Continuous monitoring or provocative testing is sometimes needed to catch it.

Persistent ST Elevation After a Heart Attack

Most ST elevation resolves within hours to days after a heart attack is treated. When it does not, persistent elevation in the same leads where the original infarction occurred raises the possibility of a left ventricular aneurysm, a thinned-out, non-contracting pouch of scar tissue on the heart wall. Studies have found that ST elevation is present in 84% to 100% of patients with an anatomically confirmed ventricular aneurysm, giving it high sensitivity as a screening clue.²³ The specificity is low, however, because LVH, bundle branch block, and normal variants can also produce chronic elevation. The distinction usually requires an echocardiogram or other imaging to confirm whether an aneurysm is actually present. But if you are looking at an ECG with old Q waves and persistent ST elevation weeks or months after a known heart attack, aneurysm should be on the list.