Transition Logo
Join Us

HRV-guided training calculator

Paste your daily HRV readings to see your own normal range, where this week sits in it, and whether the rule tested in HRV-guided training trials would keep today’s hard session.

One reading per day, ending with today. Separate days with spaces, commas or new lines, and type - or x for a day without a reading rather than leaving it blank. Whole or one-decimal milliseconds from 5 to 300, with a point for decimals, all from the same device. Up to 120 days; the most recent 60 are used.

42 days, 41 readings, 1 missed.

The example is six weeks of invented overnight readings, not a real athlete. Nothing you type leaves this page.

Below your normal range. Trial rule: swap moderate and hard work for low-intensity aerobic work or rest.

From 41 readings over the last 42 days

Below your normal range

Trial rule: swap moderate and hard work for low-intensity aerobic work or rest.

Your 7-day average of 50.5 ms is under the lower limit of 54.6 ms. In the Vesterinen and Javaloyes trials, a 7-day average outside the range replaced moderate and hard sessions with low-intensity training or rest.

7-day average
50.5 ms
6 readings in the last 7 days
Normal range
54.6 ms to 59.4 ms
Baseline 57.0 ms from 41 readings
Distance from baseline
1.42 SD below your baseline
The range is ±0.5 SD
Today's single reading
58.0 ms
On its own, inside the range

Today's reading on its own (58.0 ms) is inside the range. The trials that used this rule acted on the 7-day average rather than a single morning, which is why the verdict follows the week.

If you feel ill, this rule does not apply. HRV cannot tell you whether you are ill or whether it is safe to train.

Your last 42 days (HRV in ms)
Today's normal range (striped)Baseline (dashed)7-day averageDaily readingToday's 7-day average

How the check works

The calculator applies the session rule from the Vesterinen and Javaloyes HRV-guided training trials to a rolling baseline of the kind the Australian Institute of Sport (AIS) uses to monitor athletes. It first takes the natural log of each reading. The AIS’s 2023 best-practice guideline recommends the log of RMSSD because it makes the measurement error more uniform between high and low readings.

Your normal range is the average of those logged readings over the last 60 days, plus and minus 0.5 of their standard deviation (SD). Trials by Vesterinen and colleagues (2016, using RMSSD) and Javaloyes and colleagues (2019 and 2020, using its log) put the same width around a baseline of two to four weeks, recalculated once partway through. The AIS dashboard puts it around a rolling 60-day average, written as half the coefficient of variation. None of these trials used a continuously rolling 60-day baseline, which follows your HRV down during a long low patch (more on that below). The range needs at least 14 readings before the calculator will draw it.

The figure the rule acts on is the 7-day average: today and the six days before it, averaged on the same log scale, from at least 3 readings. Plews and colleagues found 3 readings a week to be the minimum for a valid weekly average in trained triathletes, and the AIS guideline adopts the same minimum. Outside the range, on either side, the Vesterinen and Javaloyes trials replaced moderate and hard sessions with low-intensity training or rest. Inside it they trained as planned (the calculator counts a value exactly on a limit as inside), with one difference the calculator flags: after a spell outside the range, Javaloyes and colleagues resumed hard sessions once the average was back inside, while Vesterinen and colleagues waited until it had returned to the baseline mean.

Results are shown back in milliseconds. Because the averaging happens on the log scale, the 7-day figure is a geometric mean, which sits a little below the plain average of the same numbers whenever they vary.

A worked example

The readings loaded when the page opens are six weeks of invented overnight values: five ordinary weeks between 50 and 66 ms, then a week of 49, 52, 48, a missed night, 50, 47, 58. From 41 readings the baseline is 57.0 ms and the normal range runs from 54.6 ms to 59.4 ms.

The 7-day average is 50.5 ms, 1.42 SD below your baseline, so the verdict is below the range: if today was meant to be intervals, the rule turns it into easy aerobic work or rest. Today’s single reading of 58.0 ms is inside the range on its own. A rule that looked only at this morning would clear the hard session. The 7-day rule does not, because five of the six readings this week sit below the range. The AIS guideline gives the reason for preferring the average: 7-day rolling HRV is more sensitive to training-induced change because it damps the natural day-to-day swings.

Notice too that the low week is part of the 60-day baseline. A rolling baseline moves with you, which is mostly what you want, but it means a long low patch slowly drags the range down with it. The section on getting readings you can compare comes back to this.

What the HRV-guided training trials found

Vesterinen and colleagues (2016) followed recreational runners through an eight-week block, with 31 of 40 completing it. Each morning they recorded four minutes of RMSSD lying down and compared the 7-day average with a range of mean ± 0.5 SD from four baseline weeks. Inside the range they did moderate and hard sessions; outside it, low-intensity running or rest, until the average “returned to the mean level” of the range. The HRV group ended up doing fewer moderate and hard sessions (13.2 ± 6.0 against 17.7 ± 2.5 in the group following a fixed plan). Their 3000 m running speed improved 2.1% while the fixed-plan group’s 1.1% was not significant, a small difference between groups. VO2max improved in both.

Javaloyes and colleagues (2019) used the same range and the same trigger with 17 well-trained road cyclists, measuring the log of RMSSD for 90 seconds on waking. The HRV-guided group improved peak power by 5.1%, power at the second ventilatory threshold by 13.9% and a 40-minute time trial by 7.3%, while the traditionally periodized group did not improve significantly. The direct comparison between the two groups was not statistically significant. A 2020 follow-up by the same group set the rule against block periodization in 20 cyclists; the HRV-guided group improved more measures, but fitness and performance were similar between groups at the end. An earlier study by Kiviniemi and colleagues (2007) used a different rule: a single morning reading against a 10-day reference, acting only on a drop (or a falling trend over two days), so an increase meant a hard session. Maximal running velocity rose more in that HRV group (0.9 against 0.5 km/h).

Pooled, the effect is modest. A 2021 meta-analysis by Düking and colleagues (8 studies, 198 participants) found that most HRV-guided groups did fewer moderate and hard sessions, with what the authors call a medium effect on submaximal measures (g = 0.30) but small, non-significant effects on performance and VO2peak, and fewer non-responders. Manresa-Rocamora and colleagues (2021) reached the same shape of answer: if HRV-guided training is better, “it is only by a small margin”, with less likelihood of negative responses. On average, then, the advantage over a fixed plan is small. In most of the trials, athletes following an HRV rule did fewer moderate and hard sessions without ending up less fit than athletes on a fixed plan.

Readings above the range

The Vesterinen and Javaloyes trials treated a 7-day average above the range like one below it: low-intensity work or rest. Not every protocol does. The published protocol for a later trial in professional runners (Carrasco-Poyatos and colleagues) changed training only when the 7-day average fell below the range, and Kiviniemi’s earlier single-morning rule treated a rise as a green light. Vesterinen and colleagues described either side as a sign of an abnormal situation and linked a rise to possible functional overreaching, while noting it is unknown whether HRV should be expected to rise or fall when an athlete overreaches.

A rise is harder to read than a drop. Buchheit (2014) describes HRV going up within a day of heavy training in the heat even as athletes felt worse, partly because plasma volume expands. Plews and colleagues (2013) note that in elite athletes both increases and decreases in HRV have been linked to poor adaptation, and that increases can also track gains in fitness. The calculator gives the trials’ answer. If you feel fresh and your resting heart rate is normal, keeping the session is a judgment call the trials did not test.

Reading the number in context

HRV responds to much more than training. In 9 million morning measurements from 28,175 HRV4Training users, Altini and Plews (2021) found rMSSD about 3% lower after high-intensity training and 1.6% higher after low-intensity days, 12% lower after heavy drinking and 10% lower during sickness, with a 3.2% difference between menstrual cycle phases. They call HRV a sensitive but not specific marker of stress. (Altini develops HRV4Training.) A below-range week after a late-night work deadline means something different from the same week in the middle of a hard training block, and only you know which one you are in.

Resting heart rate helps separate the cases. Plews and colleagues (2013) describe HRV saturation in elite athletes: HRV falling even though resting heart rate is also falling, which is not the same as fatigue. Buchheit (2014) suggests comparing the log of RMSSD with the RR interval to tell saturation from fatigue. If your resting heart rate has risen alongside the drop, saturation is not the explanation.

If you feel ill, this is not a training-rule question. Illness lowers HRV, but HRV cannot tell you whether you are ill or whether it is safe to train. The trials studied healthy athletes, and their rule was never meant to make that call.

Getting readings you can compare

Use one device and one kind of reading for the whole series. The AIS guideline treats on-waking and overnight measurements as equally valid and recommends whichever you can record most consistently. Garmin watches analyse wrist heart rate while you sleep and need three weeks of sleep data before showing HRV Status. Oura measures only during sleep and reports the mean of five-minute samples across the night. WHOOP reports recovery HRV as RMSSD. Apple Health stores HRV as SDNN, a different statistic, recorded automatically by Apple Watch. The trials measured RMSSD on waking. The AIS guideline’s view that overnight readings are equally valid makes an overnight RMSSD average a reasonable stand-in, though it notes they may be a little less sensitive to training than waking readings. No trial used Apple’s SDNN, so treat a verdict built from it as rough. Never mix sources: values from different devices or statistics sit on different scales. Garmin’s HRV Status already compares a seven-day average with a baseline range; what this page adds is a published width, the trials’ session rule and the return-to-baseline check, with every number shown.

Mark nights without a reading with - or x so the 7-day window lines up with the calendar. 3 readings in 7 days is the minimum; Buchheit suggests at least 3 to 4 a week, and more gives a steadier average. The trials built their ranges from four weeks of baseline readings, so treat a range from the minimum 14 readings as provisional. Because the averaging is done on the log scale, one implausibly low night weighs heavily: with a few weeks of steady history, a single 20 ms reading in a week of 60s is usually enough to drag the 7-day average below the range, and it widens the range itself for as long as it stays in the 60-day window. If a reading looks like a sensor glitch, replace it with - rather than average it in.

A rolling range keeps up with your fitness, but it can drift. Nuuttila and colleagues (2022), who also updated their limits continuously, saw some individuals’ limits slowly slip downward (their example was a heart-rate-to-running-speed index rather than HRV) and suggested checking a short-term and a long-term limit together, such as 2 and 8 weeks. If your range now sits well below where it was during a settled month, give an inside verdict less weight.

What this check cannot tell you

The width of 0.5 SD is a convention, not a measured threshold. Buchheit (2014) points out there is no evidence that changes beyond any particular fraction of an athlete’s normal variation are meaningful in practice, and Manresa-Rocamora and colleagues found it remains unclear whether a single-day or rolling-average rule works better. The trials were small (no more than 40 athletes each), short, used baselines of a few weeks updated at most weekly rather than a rolling 60-day one, and none studied triathletes training three sports at once. The AIS guideline uses this kind of range to flag a meaningful change that then has to be interpreted, and it recommends ignoring the built-in training verdicts some wearables give. So the verdict here is what the trial rule would have prescribed today, not an instruction, and it does not predict how the session will go. A good day on a low reading is entirely possible.

Where Transition fits

Transition’s Recovery Intelligence makes a related comparison each day from connected health sources such as Garmin, Oura, WHOOP and Apple Health. It compares your 7-night HRV average and 7-night resting heart rate average with your own 60-day baselines and flags an HRV average more than half your usual variability below baseline, or a resting heart rate that far above it. It needs 14 nights of history and uses recent sleep as a check. Unlike this calculator, it does not flag a high HRV average, and it weighs resting heart rate alongside HRV.

Common questions

What is a normal HRV range?
Normal HRV is personal. Across 28,175 HRV4Training users (Altini and Plews 2021), morning rMSSD averaged 67 ± 34 ms for men and 68 ± 34 ms for women, so population charts say little about you. One common convention, used by the Australian Institute of Sport dashboard and in the Vesterinen and Javaloyes training trials, calls the mean of your own readings plus or minus 0.5 standard deviations your normal range, worked out on the natural log of the readings as the AIS recommends. The AIS dashboard uses the last 60 days, and this calculator needs at least 14 readings from that window before it draws one.
Should I do a hard workout if my HRV is low?
Not on one low morning, under the 7-day rule used in the Vesterinen (2016) and Javaloyes (2019) trials. Athletes compared their average HRV over the last 7 days with a personal range of the baseline mean ± 0.5 standard deviations, built from four weeks of morning readings. When that 7-day average fell below or rose above the range, moderate and hard sessions were replaced with low-intensity training or rest; inside the range they trained as planned, although Vesterinen's runners waited until the average had climbed back to the baseline mean after a low spell. An older trial acted on single mornings instead, and which approach works better has not been tested head to head. This calculator applies the 7-day comparison against a rolling 60-day range and needs at least 3 readings in the last 7 days, the weekly minimum Plews and colleagues found valid. If you feel ill, treat it as a health question rather than a training decision.
How many days of HRV data do I need?
For this calculator, at least 14 readings within the last 60 days to set a normal range, and at least 3 readings within the last 7 days for the 7-day average. The trials behind the rule built their ranges from four weeks of morning readings, and Plews and colleagues found 3 readings a week to be the minimum for a valid weekly average in trained triathletes.
What does it mean if my HRV is above my normal range?
In the Vesterinen (2016) and Javaloyes (2019) trials, which compared a 7-day average with a range of the baseline mean ± 0.5 standard deviations, a 7-day average above the range was treated the same as one below it, with low-intensity training or rest instead of moderate or hard sessions. Not every protocol does this: a later protocol in professional runners changed training only on drops, and an older single-morning rule (Kiviniemi 2007) treated a rise as a green light. A rise can come with coping well, but it has also been seen after heavy training in the heat and is discussed as a possible sign of overreaching, so it is harder to interpret than a drop.
Can I use Apple Watch HRV readings?
You can, as a rough guide only. Apple Health stores HRV as SDNN, recorded automatically by Apple Watch, while the trials behind this rule measured RMSSD each morning, and no trial has tested the rule on SDNN. Use one device and one kind of reading for the whole series and never mix Apple Watch SDNN values with RMSSD values from another device, because they sit on different scales.
Does HRV-guided training make you fitter?
Only slightly more than a fixed plan, on average. A 2021 meta-analysis of 8 studies with 198 participants found what the authors call a medium effect on submaximal measures (g = 0.30) but small, non-significant effects on performance and VO2peak, with fewer non-responders; a second 2021 meta-analysis concluded any advantage is “only by a small margin”. In most trials, HRV-guided groups did fewer moderate and hard sessions than groups following a fixed plan without ending up less fit.

Sources