The R-R interval is the time between two consecutive R-peaks on an ECG tracing, and you measure it by counting the small squares (or using calipers) between one R-wave peak and the next. On standard ECG paper running at 25 mm per second, each small square equals 0.04 seconds (40 milliseconds). Count the squares between two R-peaks, multiply by 40, and you have the interval in milliseconds.
Reading R-R Intervals on Paper ECG
Standard ECG paper moves through the machine at 25 mm per second. The grid is divided into small squares (1 mm each) and large squares (5 mm each, outlined by thicker lines). Each small square represents 40 ms of time, and each large square represents 200 ms. To measure an R-R interval, locate the tallest, sharpest upward spike in the QRS complex. That peak is the R-wave. Find the same peak in the very next beat, then count the number of small squares between them.
If you count 20 small squares between two R-peaks, the interval is 20 × 40 = 800 ms. For greater precision, use ECG calipers. Place one point of the caliper on an R-peak, spread the other point to the next R-peak, then lay the caliper against the grid to read the distance. In clinical validation studies, calipers typically measure to the nearest 0.25 mm, which translates to 10 ms of resolution. Measuring three consecutive intervals and averaging them gives a more reliable value than a single measurement, especially when rhythm is slightly irregular.
Converting to Heart Rate
Heart rate and R-R interval have a simple inverse relationship: multiply them together and you always get 60,000. So to convert an R-R interval in milliseconds to beats per minute, divide 60,000 by the interval. An R-R interval of 800 ms gives you 60,000 ÷ 800 = 75 beats per minute. Going the other direction, if you know heart rate is 60 bpm, the R-R interval is 60,000 ÷ 60 = 1,000 ms, or exactly one second.
One thing to keep in mind: this relationship is hyperbolic, not linear. That means a change of 100 ms at a fast heart rate shifts bpm much more than the same 100 ms change at a slow heart rate. This is one reason researchers often prefer working in milliseconds rather than bpm when analyzing heart rhythm variability.
Normal R-R Interval Range
In a study of 174 healthy people aged 16 to 89, the mean resting R-R interval was about 865 ms, with a normal range spanning roughly 655 to 1,141 ms. That corresponds to heart rates between about 53 and 92 bpm. The mean interval did not differ significantly by sex, age, or smoking status, although the amount of beat-to-beat variation in interval length did decrease with age. If your measured interval falls within that range during rest, the underlying rate is considered normal for a healthy adult.
Measuring With Digital Devices
Most digital ECG systems and wearable heart monitors detect R-peaks automatically using signal processing algorithms. The general approach involves filtering the raw electrical signal to remove background noise and power line interference, then amplifying the steep slopes unique to the QRS complex so they stand out from slower P and T waves, and finally applying a threshold to mark each R-peak’s exact location. Once peaks are identified, the software simply subtracts one timestamp from the next to produce R-R intervals.
Sampling rate matters. A device recording at 250 Hz or higher produces R-R intervals accurate enough for any type of heart rate variability analysis. At 100 Hz, basic time-based measurements remain acceptable, but frequency-based analyses (which break variability into different rhythm bands) lose precision. Below 50 Hz, measurements become unreliable for any serious analysis. If you’re choosing a device specifically for R-R interval work, look for at least 250 Hz sampling.
Optical Sensors vs. ECG
Smartwatches and fitness trackers typically use an optical pulse sensor (photoplethysmography) on the wrist rather than electrical leads on the chest. Instead of detecting the R-wave directly, these sensors detect the pulse of blood arriving at the wrist a fraction of a second later. Despite this indirect approach, comparison studies show extremely high agreement between optical and ECG-derived intervals, with correlation values above 0.95 for all standard variability metrics. Average R-R intervals from optical sensors matched ECG values almost exactly (718.9 ms vs. 718.8 ms in one controlled study). The catch is that accuracy drops during vigorous movement or if the sensor fits loosely, so wrist-based measurements are most reliable at rest.
Cleaning Up Your Data
Raw R-R interval series almost always contain artifacts. Some are technical, like electrical noise causing a false peak detection or a missed beat. Others are physiological, like premature heartbeats (ectopic beats) that fire earlier than expected and create one abnormally short interval followed by one abnormally long interval. Both types distort any analysis you run on the data.
The simplest cleaning method is threshold filtering: flag any interval that differs from the surrounding intervals by more than a set percentage (commonly 20 to 25%) and either remove it or replace it with an interpolated value. More sophisticated approaches use adaptive filters that learn the expected pattern from the surrounding rhythm and correct only the intervals that deviate from it. If you’re doing this manually in a spreadsheet, scanning for sudden jumps or drops in the interval series and replacing them with the average of the neighboring intervals is a reasonable starting point. For research-quality analysis, automated preprocessing tools that combine noise removal with ectopic beat detection give more consistent results.
What R-R Intervals Tell You Beyond Heart Rate
The real power of R-R interval measurement comes from analyzing how much the intervals vary from beat to beat. This is heart rate variability, or HRV, and it reflects how actively your nervous system adjusts your heart’s rhythm. Two key metrics dominate:
- SDNN: The standard deviation of all normal R-R intervals over a recording period, measured in milliseconds. It captures overall variability from all sources and is considered the gold standard for cardiac risk assessment when calculated from a full 24-hour recording. Both the “fight or flight” and “rest and digest” branches of the nervous system contribute to SDNN, though in shorter recordings, the rest-and-digest branch dominates.
- RMSSD: The root mean square of successive differences between consecutive R-R intervals. You calculate it by finding the difference between each pair of adjacent intervals, squaring those differences, averaging the squares, and taking the square root. RMSSD captures rapid, beat-to-beat fluctuations and primarily reflects the calming influence of the vagus nerve on the heart. It’s the metric most wearable devices report as your “HRV score.”
Both metrics start with the same raw material: a clean series of R-R intervals measured in milliseconds. The more carefully you measure and clean those intervals, the more meaningful the variability analysis becomes. For RMSSD in particular, even small timing errors compound quickly because the calculation depends on differences between adjacent intervals rather than their absolute values. This is why the 250 Hz minimum sampling rate recommendation exists: at lower rates, the tiny timing imprecisions in each interval measurement add up and inflate or deflate the variability estimate.