How to Read an ECG: A Step-by-Step Guide

An electrocardiogram (ECG, also called an EKG) records the heart’s electrical activity through electrodes on the skin. Learning to read one can help you understand the tracing and communicate observations, but interpretation takes training and must be combined with symptoms, examination, and other tests. Do not use this guide to diagnose yourself or delay care.

If someone has chest pressure, severe shortness of breath, fainting, or another potentially urgent symptom, seek emergency medical care rather than trying to interpret an ECG. A normal-looking tracing does not rule out every serious problem.

What You’ll Need

A complete 12-lead ECG tracing, with the patient’s age and relevant clinical context when available.
The printed calibration information or device settings. A common standard is 25 mm per second and 10 mm per millivolt, but verify the tracing rather than assuming.
A ruler or the ECG’s grid, plus a reference for normal ranges and a qualified clinician for interpretation.

Step-by-Step

Confirm the tracing. Check that the name or identifier and recording date belong to the right person. See whether all 12 leads are present and the tracing is labeled correctly. Compare with earlier ECGs if a clinician has them; changes over time can matter.
Check calibration and signal quality. Read the paper speed and voltage scale. At the common settings, one small horizontal box represents 0.04 seconds and one large box 0.20 seconds; vertically, 10 millimeters represents 1 millivolt. Look for baseline drift, muscle tremor, loose-electrode artifacts, or clipped waves. If quality is poor, request a clean recording before drawing conclusions.
Find the heart rate. For a regular rhythm at 25 mm/s, divide 300 by the number of large boxes between consecutive R waves. Alternatively, count QRS complexes in a 10-second strip and multiply by six. These are estimates; the second method is more useful when the rhythm is irregular. Compare the result with the person’s situation and the device’s stated rate.
Assess rhythm regularity. Compare the spacing between successive R waves across the strip. Then look for a consistent P wave before each QRS and a consistent relationship between P waves and QRS complexes. A regular pattern supports a regular rhythm, but irregular patterns require careful review; do not label a rhythm from one feature alone.
Inspect P waves and atrial activity. P waves represent electrical activation of the atria. Check whether their shape and direction are consistent in the leads where they are normally expected, and whether each appears to conduct to a QRS. Missing, unusual, or multiple atrial deflections can have several explanations and need professional interpretation.
Measure the PR interval. Measure from the start of the P wave to the start of the QRS. At standard speed, count small boxes and multiply by 0.04 seconds. A typical adult PR interval is about 0.12–0.20 seconds, but age, rhythm, and clinical setting matter. A value outside a reference range is a finding to assess, not a diagnosis by itself.
Examine the QRS complex. Measure from its first deflection to its end. A typical adult QRS is under 0.12 seconds; a wider complex can arise for different reasons and merits clinical review. Look for consistent shape across beats and leads, and note whether the computer interpretation flags a possible abnormality without treating that label as final.
Estimate the electrical axis. The axis describes the overall direction of ventricular electrical activity in the frontal plane. For a basic screen, compare QRS direction in leads I and aVF: both mostly upward suggests a broadly normal direction; opposite directions may suggest axis deviation. This shortcut is only an estimate and is less reliable with unusual waveforms or technical issues.
Review ST segments and T waves. Compare these portions across neighboring leads, looking for elevation, depression, or an unusual T-wave shape. Their meaning depends on the lead, baseline, symptoms, prior tracings, and other findings. Do not apply a single cutoff or interpret an isolated change as proof of a heart attack.
Summarize and seek interpretation. Record objective observations in order: rate, regularity, P-to-QRS relationship, intervals, QRS appearance, axis estimate, and ST–T features. Include symptoms and relevant history when sharing the tracing with a clinician. A trained professional should make the interpretation and decide whether further assessment is needed.

Common Mistakes

Skipping the setup: Incorrect calibration, reversed leads, or poor contact can distort the tracing. Confirm settings and signal quality first.
Using a regular-rhythm shortcut on an irregular tracing: The 300-box method assumes fairly even R–R spacing. Use a longer strip or an appropriate average instead.
Trusting automated labels as diagnoses: ECG software can help identify patterns, but artifacts and unusual circumstances can mislead it. The tracing needs clinical review.
Reading one lead in isolation: The same electrical event appears differently across leads. Assess related leads and the full tracing.
Assuming normal means safe: Some conditions are intermittent or may not appear on a single ECG. Symptoms and clinical assessment still matter.
Confusing observation with diagnosis: Describe what is visible rather than naming a condition unless qualified to interpret it.

Tips

Use the same sequence every time; a checklist reduces skipped steps.
Mark the beginning and end of intervals before counting grid boxes.
Compare with an earlier ECG only when patient identity, lead placement, and calibration are appropriate.
Separate artifact from a true change by checking whether it appears consistently across beats and neighboring leads.
Keep the tracing with its clinical context. Rate and waveform findings mean different things in different situations.

Troubleshooting

The waves are hard to see: Check the scale, paper speed, contrast, and baseline. Ask for a clearer recording if there is movement, electrical interference, or a loose electrode; do not force an interpretation from a poor-quality tracing.

The rate methods disagree: The rhythm may be irregular, the strip may be too short, or the R waves may be difficult to identify. Count over a longer interval and have a clinician verify the rate.

The machine reports an abnormality: Treat it as a prompt for review, not a confirmed diagnosis. Share the complete ECG and relevant symptoms with a healthcare professional.

The tracing looks normal but symptoms continue: Do not use the ECG to rule out a problem. Contact a healthcare professional; seek emergency care for severe or sudden symptoms.

Frequently Asked Questions

What is the difference between an ECG and an EKG?

They are two abbreviations for the same test, an electrocardiogram. ECG comes from the English spelling; EKG reflects the German spelling.

Can I read my own ECG?

You can learn to recognize basic parts of the tracing, but self-interpretation cannot establish a diagnosis. Ask a qualified clinician to explain your result in context.

What do the 12 leads represent?

They are different electrical views of the heart, recorded using electrodes placed on the limbs and chest. Together they provide more information than any single lead.

Does a normal ECG rule out a heart problem?

No. A single ECG is a snapshot, and some problems may be intermittent or not visible on that recording. Symptoms and other clinical information remain important.

Why might an ECG look different from one recording to another?

Heart rate, body position, electrode placement, movement, and changing health conditions can affect the tracing. A clinician can determine whether a difference is technical or meaningful.

Further reading: MedlinePlus. Details can change, so check the official page for anything that matters to you. Spotted a mistake? Tell us. How we write an answer.