ECG - JSON Export Format
Documentation for ECG structure in JSON exports
Last updated: October 31, 2026
JSON exports store ECG recordings in the ecg array.
Structure
This example shows a 30-second recording. The voltageMeasurements array is shortened to its first two readings; numberOfVoltageMeasurements describes the original recording.
{
"start": "2024-02-06 14:30:00 -0800",
"end": "2024-02-06 14:30:30 -0800",
"classification": "Sinus Rhythm",
"averageHeartRate": 72,
"numberOfVoltageMeasurements": 15360,
"voltageMeasurements": [
{"date": 1707258600, "voltage": 123, "units": "mcV"},
{"date": 1707258600.001953125, "voltage": 125, "units": "mcV"}
],
"samplingFrequency": 512,
"source": "Apple Watch"
}
Fields
start,end(String): Recording timestamps inyyyy-MM-dd HH:mm:ss Zformat.classification(String): The recorded classification. English values emitted by the exporter areNot Set,Sinus Rhythm,Atrial Fibrillation,Inconclusive Low Heart Rate,Inconclusive High Heart Rate,Inconclusive Poor Recording,Inconclusive, andUnrecognized.averageHeartRate(Number): Average heart rate in beats per minute.numberOfVoltageMeasurements(Number): Total voltage-reading count in the recording.voltageMeasurements(Array): Recorded voltage readings.samplingFrequency(Number): Sampling frequency in hertz.source(String): The recording app or device.
Voltage measurements
Each reading contains:
date(Number): Measurement time as Unix epoch seconds, including fractional seconds.voltage(Number): Voltage in microvolts.units(String):mcV.
ECG exports can be large because they include the voltage series. Use a shorter export date range when you need smaller files.