HRV on your running watch: how to use it in training

What watch HRV is, whether the wrist sensor is reliable and when to swap a hard session for an easy one, with what the research shows and a decision table.

By Coach and creator of Magic Training

18 min read

Row of Polar heart rate watches on colorful acrylic stands, displayed on a white counter at a sports expo
In this article

HRV (heart rate variability) is the beat-to-beat variation in the time between heartbeats, and most watches measure it during sleep. The number is one more signal about your recovery, not a training order. Compare your last 7 days with your personal range, never a single night, and check it against sleep, perceived effort and resting heart rate. In meta-analyses of studies that used HRV to set training, the advantage over a fixed plan in VO2max and performance was small and not statistically significant, with fewer hard sessions and similar results.

Here you will see what the watch is measuring, how each brand builds its "status", how much the wrist sensor gets wrong, what the research shows and how to turn the number into a decision.

Key takeaways

  • Watch HRV is the variation between heartbeats, measured on most devices during sleep. COROS and WHOOP work with rMSSD and Apple uses SDNN; Garmin and Polar do not name the index on the pages consulted. So do not compare numbers across brands.
  • The wrist sensor agrees unevenly with the electrocardiogram, depending on the model. Across 536 nights in 13 adults, the mean error in rMSSD ran from 5.96% (Oura Gen 4) to 16.32% (Polar Grit X Pro), with 10.52% for the Garmin Fenix 6.
  • A single night swings a lot. The typical error of resting Ln rMSSD is close to 12%, and the weekly average tracked performance better than a single day's value.
  • In the Manresa-Rocamora (2021) meta-analysis of HRV-guided training, the effect on VO2max (SMD 0.20) and performance (SMD 0.20) was small and not significant. A second meta-analysis (Düking, 2021) found fewer non-responders and more positive responders in performance.
  • In practice: a 7-day average below your range, plus poor sleep or high effort at the same pace, justifies swapping the quality session for an easy run. It is not a reason to skip the week.

What is HRV: SDNN and rMSSD

The heart does not beat like a metronome. The interval between two beats changes from one to the next, and HRV quantifies that change. The 1996 standards document from the Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology defined how to measure and interpret HRV. It is a widely cited reference in HRV studies (Task Force, 1996).

Two indices appear on watches. SDNN is the standard deviation of the normal intervals between beats. rMSSD is the square root of the mean of the squared differences between successive intervals. A 2017 review describes rMSSD as the main time-domain measure for estimating vagally mediated changes. It states that rMSSD is more influenced by the parasympathetic system than SDNN is (Shaffer and Ginsberg, 2017).

In watch terms, HRV above your normal suggests more parasympathetic activity, the rest-and-digest side, as the Polar page puts it (Polar, Nightly Recharge). Higher does not mean "better" in general. In elite athletes, Plews and colleagues describe "saturation": HRV falls even while resting heart rate also falls, and fitness can improve with an atypical drop in HRV (Plews et al., 2013a).

How your watch builds its status number

Watches measure at the wrist, during sleep or while you are still, and compare the result with your own history. What changes is the index, the comparison window and the final label. The table summarizes what the official pages say, and nothing more.

BrandWhen it measuresIndexComparison with youLabel
Garmin (HRV Status)during sleep; recommends wearing the watch to bednot named on the pages consulted7-day average against your personal range; needs 3 weeks of dataBalanced, Unbalanced, Low, Poor (Poor = well below normal for your age)
Apple (Health/Heart Rate)automatic samples; "Recovery HRV" is measured while you are stillSDNN, according to the HealthKit documentationthe pages consulted do not describe a baseline rangevalue in ms, no label on those pages
Polar (Nightly Recharge)about the first 4 hours of sleepHRV, together with heart rate and breathing, in a single valueagainst your usual level over the last 28 daysANS charge from -10 to +10; near zero is your normal
COROSduring sleep; nightly averagerMSSD (the page text writes "mRSSD" and the chart writes rMSSD)a "normal range" that appears after a few daysLower, Within normal range, Higher
WHOOPRecovery is calculated on waking; the page does not say what part of the night HRV is measuredrMSSD (field hrv_rmssd_milli in the API)the page does not describe weighting or windowRecovery from 0 to 100% with HRV and resting HR as inputs

Two practical consequences. First, SDNN and rMSSD are different measures, so 45 ms on Apple and 45 ms on COROS do not say the same thing. Second, the range is yours. Both Garmin and COROS say you should not compare your status with someone else's, because the normal range differs from one runner to another.

Can you trust the wrist sensor?

In part: the evidence is stronger for the average of several days than for a single night's value, and the absolute error depends on the model. Validations compare the watch with an electrocardiogram (ECG). They show error that depends on the model, the position, the time of day and how the manufacturer calculates.

The most relevant study for your case measured 536 nights in 13 healthy adults with ECG and five wrist and finger devices. For rMSSD, the mean absolute percentage error (MAPE) was 5.96% for the Oura Gen 4, 7.15% for the Oura Gen 3, 8.17% for the WHOOP 4.0, 10.52% for the Garmin Fenix 6 and 16.32% for the Polar Grit X Pro. The authors rated Garmin and Polar as poor agreement, with Lin's concordance coefficients of 0.87 and 0.82 (Dial et al., 2025).

Another study, with 53 adults during one night in a lab, compared six devices with ECG. The intraclass correlation for HRV was 0.67 on the Apple Watch Series 6 (underestimating by 9.6 ms on average). It was 0.24 on the Garmin Forerunner 245 Music, whose HRV came from a manual 3-minute test while awake. It was 0.99 on the WHOOP 3.0, the only device that delivered the raw R-R intervals. The authors note that comparisons with proprietary algorithms are "naive" (Miller et al., 2022).

For training, the central point is consistency. In recreational runners, nocturnal HRV (Ln rMSSD) was highly reproducible between two nights after identical easy sessions, with an ICC of 0.92 to 0.97. It dropped on the night after a maximal 3,000 m test (Nuuttila et al., 2022).

The validation studies above used healthy young adults and specific models, so the results do not apply to every watch. That favors following the trend on the same device, not a single night's value or a comparison between brands.

How much the number swings from day to day

Daily swings are large and depend on the person, so one night alone says little. Buchheit lists a typical error, expressed as a coefficient of variation, of about 12% for resting Ln rMSSD. He also describes changes of 10 to 20% associated with daily load changes, and intense exercise lowers vagal indices for 24 to 48 hours (Buchheit, 2014). As an order of magnitude, a 10% drop in one night fits within that noise.

The average solves much of the problem. In 10 runners who responded to 9 weeks of training, the change in 10 km time had a very large correlation with the week's mean Ln rMSSD (r = -0.76) and a small one with a single day's value (r = -0.17) (Plews et al., 2013b). The group recommends at least 3 valid measurements per week (Plews et al., 2014).

In two elite triathletes followed for 77 days, the 7-day rolling average of Ln rMSSD fell toward the event in the athlete who ended up diagnosed with non-functional overreaching. It stayed stable in the other. The coefficient of variation of the rolling average also fell in him (Plews et al., 2012).

This is a case study of athletes training 23 hours a week. It works as an indication that the trend matters, not as a rule for people who train five or six days.

What research says about HRV-guided training

HRV-guided training worked at least as well as a fixed plan in most studies, with fewer moderate or hard sessions, but the advantage in VO2max and performance is small and uncertain. In the two trials below, HRV was measured every morning and set the day's training. Within the participant's individual range, training was moderate or hard; outside it, training was easy (or rest, in the Kiviniemi study).

StudyWhoResult as published
Kiviniemi and colleagues, 200726 moderately fit men; 4 weeksHRV group: peak VO2 from 56 to 60 ml/kg/min; maximal speed from 15.5 to 16.4 km/h. Predefined group: peak VO2 from 54 to 55 (no significant change); maximal speed from 15.1 to 15.7. Between-group difference in maximal speed: P = 0.048. No significant difference in peak VO2
Vesterinen and colleagues, 201640 recreational runners; 8 weeks of intense trainingHRV group did fewer moderate and hard sessions (13.2 versus 17.7). Over 3,000 m, the HRV group improved 2.1% (significant within the group) and the predefined group 1.1% (not significant); small between-group effect: 0.42. VO2max rose in both: 3.7% in the HRV group and 5.0% in the predefined group

A systematic review and two meta-analyses pooled these studies. The review by Düking and colleagues (2020), running only, found 5 studies and 166 participants. In four interventions the HRV group reduced moderate and hard training, and in five the performance gain was larger with HRV. Peak VO2 improved in both models, with no clear difference. The authors call the findings limited and inconsistent.

The meta-analysis by Manresa-Rocamora and colleagues (2021) estimated standardized differences favoring HRV of 0.50 in vagal indices (95% CI 0.09 to 0.91). For the other outcomes it found only small differences:

  • VO2max: 0.20 (CI -0.07 to 0.47)
  • second ventilatory threshold: 0.26 (CI -0.05 to 0.57)
  • endurance performance: 0.20 (CI -0.09 to 0.48)

The authors conclude that, if the advantage in fitness and performance exists, it is by a small margin, with a lower chance of negative responses.

A second meta-analysis, by Düking and colleagues (2021), pooled 8 studies and 198 participants. It found a positive effect, which the authors rate as medium, on submaximal parameters (g = 0.296; CI 0.031 to 0.562). Effects on performance (g = 0.079; p = 0.597) and peak VO2 (g = 0.171; p = 0.130) were small and not significant. In performance, the HRV group had fewer non-responders and more positive responders.

How to read this honestly: there are few studies, small samples and different protocols. Manresa-Rocamora and colleagues say that the choice of index, the recording position and the way the range is defined still need study. The two trials above used a morning measurement and their own rule, not the status label of a watch. One possible reading, but not a demonstrated one, is that the gain comes from avoiding hard sessions on bad days.

How to decide the day's training with HRV

Use HRV as a filter, not as a judge. The sequence below combines what the research and the manufacturers support with training common sense.

  1. Wait for the range to exist. Garmin asks for 3 weeks of sleep with the watch on your wrist, and Polar compares against the last 28 days. Before that there is nothing to compare with.
  2. Look at the 7-day average against the range. It is the basis of the Garmin status. A single night works as a warning, not a verdict.
  3. Cross-check with other signals. Buchheit notes that heart rate measures do not capture every aspect of well-being and fatigue, and suggests combining them with a training diary and questionnaires. Ask yourself: does your usual easy pace feel harder? Did you sleep badly? Has resting HR gone up? Is there muscle soreness? Training zones help keep easy truly easy, and the heart rate calculator gives the HR ranges.
  4. Swap the intensity, not the week. In the Kiviniemi and Vesterinen trials, the adjustment was made day by day, swapping the intensity (or resting, in Kiviniemi). As a training rule: if the quality session falls on a bad day, do an easy run and try the session the next day, if the signals return to normal.
  5. It is not a diagnosis. COROS states that the values are for reference only and do not diagnose medical conditions.

If you prefer video, Abner Santana's channel has HRV and training load: when to train hard or rest in running (video in Portuguese). It covers HRV, perceived effort and training load as a simple dashboard for deciding when to train hard and when to rest.

On Magic Training, the Performance page shows fitness, fatigue and form calculated from your workouts. With a connected watch (Garmin, COROS, Apple Health or Health Connect on Android), it also shows a readiness reading that compares your resting HR, HRV and last night's sleep with your 30-day average. When more than one signal is off its usual level, it suggests easing the intensity; if a hard workout is scheduled, it suggests swapping it for an easy day. It is a Premium feature and is not a diagnosis.

Catalog plans can be sent to your watch; see how it works on Garmin, Apple Watch and Wear OS, and see the Magic Training apps.

What you see on the watch, what it tends to mean and what to do

The table below translates the most common cases. It is a rule of thumb built on the cited sources, not a study finding, and it does not replace your coach or a doctor.

What you see on the watchWhat it tends to meanWhat to do
One night below range; 7-day average within; sleep and legs normalNoise, a worse night or a hard workout the day before. Fits within the typical measurement errorFollow the plan, warm up and see how the body responds. Check again tomorrow
One night below range after alcohol, a late workout, illness or short sleepKnown effect on the night, not a loss of fitnessDiscard the reading. If there are signs of illness, take it easy
7-day average within rangeLoad tolerated at the momentFollow the plan. Nothing needs to change
7-day average below range for several days, with poor sleep, higher resting HR or a heavy easy paceAccumulated fatigue, stress or illness coming onSwap the next quality session for an easy run and check the week's load
Average above range (Unbalanced status or higher on Garmin)May be good recovery, but Garmin says it can also mean the body is working too hard to recoverDo not count it as a green light. Follow how you feel and your sleep
Values that jump a lot from one night to the next, with no clear reasonIrregular sleep, a loose strap or a failing sensorAdjust the watch, always sleep with it on and look at the average
A drop that goes unexplained for weeks, or low HRV with palpitations, chest pain, dizziness or faintingBeyond what the watch can interpretSee a doctor

When not to read last night's HRV

The night is affected by several things that have nothing to do with your training. The most studied is alcohol. In 4,098 people with continuous beat-to-beat recording, the HRV-derived recovery indicator in the first 3 hours of sleep fell on average by 9.3 percentage points with a low dose (up to 0.25 g/kg), 24.0 with a moderate dose and 39.2 with a high dose (Pietilä et al., 2018).

Polar lists stress, late-night exercise, illness and alcohol as factors that raise heart rate and reduce HRV early in the night. It also says the strap must be worn snug on the wrist. To give a status, it asks for more than 4 hours of sleep and three consecutive nights of successful measurement (Polar). COROS adds that a single irregular beat (arrhythmia) can distort the reading (COROS). Garmin cites poor sleep, stress, insufficient nutrition and a drink or two (Garmin).

In athletes with a long training history, Buchheit cites pre-competition stress as a factor that can reinforce the drop in vagal indices. In taper week, read the number with that in mind, and see the marathon taper guide.

Common mistakes

Reacting to a single night

This is the most frequent mistake. The night comes with measurement and routine noise, and a good decision comes from the 3 to 7 day average and a signal that confirms it. A single low reading rarely justifies changing the week.

Comparing with another person or another brand

The range is personal and the indices differ between brands. If you switch watches, the range starts again from zero, and the old 45 ms does not carry over to the new device.

Reading the night after a hard workout as lost fitness

Intense exercise lowers vagal indices for 24 to 48 hours, and the maximal 3,000 m test reduced the next night's HRV in the runners studied. A drop after intervals or the long run is the expected response. What matters is whether the average returns to range in the following days. See also post-run recovery.

Cutting the whole week because of the number

The trials swapped the day's intensity, not the week. A marathon plan has little margin for lost weeks, and the minimum duration is in how many weeks to train for a marathon. Total load also matters, and Abner Santana's channel has How to avoid running injuries: the long-run mistake and the 10% rule (video in Portuguese).

Chasing a higher number

Higher HRV is not a goal. In well-trained athletes, fitness can improve with lower HRV, and Garmin warns that values above the range can also mean the body is working too hard to recover. Train by feel, by the plan and by performance, and use the number as context.

Frequently asked questions

What is a normal HRV for a runner?

There is no universal value. It depends on age, sex, the index (SDNN or rMSSD) and the device. Garmin and COROS say the normal range is personal and that comparing with another person is not valid. Compare only with your own range, built in the first weeks of using the watch.

Can I trust the HRV on my Garmin or Apple Watch?

For the trend, yes, with caution. Against ECG, the error varies by model. In one full-night study it ran from 5.96% (Oura Gen 4) to 16.32% (Polar Grit X Pro) for rMSSD, with 10.52% for the Garmin Fenix 6; the Apple Watch was not in that study. The absolute value changes from one brand to another. Always use the same device, sleep with it snug and look at the 7-day average, not the night.

Should I cancel a workout if my HRV is low?

Not for one night. If the 7-day average is below range and there is another signal, such as poor sleep or a heavy easy pace, swap the quality session for an easy run. The guided-training trials changed the day's intensity and did not cancel the week. Symptoms such as palpitations or chest pain call for a doctor.

Is high HRV always good?

No. Garmin says a value above the range can mean the body is working too hard to recover, and Plews and colleagues describe saturation in elite athletes, in which HRV falls while fitness improves. Treat "above range" as one more data point, and confirm it with sleep and how you feel.

Is an overnight wrist reading as good as a morning chest-strap one?

They are not interchangeable. In 24 recreational runners (Nuuttila et al., 2024), nocturnal wrist-sensor and morning chest-strap measurements correlated at baseline (r = 0.42 to 0.91). They responded differently to intensified training, especially in LnRMSSD. Pick one method and stick with it, rather than mixing sources.

References

  • Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability: standards of measurement, physiological interpretation and clinical use. Circulation, 1996;93(5):1043-65. PubMed 8598068 (also published in the European Heart Journal, 1996;17(3):354-81, PubMed 8737210). Only the bibliographic record was consulted.
  • Shaffer F, Ginsberg JP. An overview of heart rate variability metrics and norms. Frontiers in Public Health, 2017;5:258. PMC5624990
  • Plews DJ, Laursen PB, Stanley J, Kilding AE, Buchheit M. Training adaptation and heart rate variability in elite endurance athletes: opening the door to effective monitoring. Sports Medicine, 2013a;43(9):773-81. PubMed 23852425
  • Plews DJ, Laursen PB, Kilding AE, Buchheit M. Evaluating training adaptation with heart-rate measures: a methodological comparison. International Journal of Sports Physiology and Performance, 2013b;8(6):688-91. PubMed 23479420
  • Plews DJ, Laursen PB, Le Meur Y, Hausswirth C, Kilding AE, Buchheit M. Monitoring training with heart rate-variability: how much compliance is needed for valid assessment? International Journal of Sports Physiology and Performance, 2014;9(5):783-90. PubMed 24334285
  • Plews DJ, Laursen PB, Kilding AE, Buchheit M. Heart rate variability in elite triathletes, is variation in variability the key to effective training? A case comparison. European Journal of Applied Physiology, 2012;112(11):3729-41. PubMed 22367011
  • Buchheit M. Monitoring training status with HR measures: do all roads lead to Rome? Frontiers in Physiology, 2014;5:73. PMC3936188
  • Miller DJ, Sargent C, Roach GD. A validation of six wearable devices for estimating sleep, heart rate and heart rate variability in healthy adults. Sensors, 2022;22(16):6317. PMC9412437
  • Dial MB, Hollander ME, Vatne EA, Emerson AM, Edwards NA, Hagen JA. Validation of nocturnal resting heart rate and heart rate variability in consumer wearables. Physiological Reports, 2025;13(16):e70527. PMC12367097
  • Nuuttila OP, Seipäjärvi S, Kyröläinen H, Nummela A. Reliability and sensitivity of nocturnal heart rate and heart-rate variability in monitoring individual responses to training load. International Journal of Sports Physiology and Performance, 2022;17(8):1296-303. PubMed 35894977
  • Nuuttila OP, Kyröläinen H, Kokkonen VP, Uusitalo A. Morning versus nocturnal heart rate and heart rate variability responses to intensified training in recreational runners. Sports Medicine - Open, 2024;10(1):120. PMC11541970
  • Kiviniemi AM, Hautala AJ, Kinnunen H, Tulppo MP. Endurance training guided individually by daily heart rate variability measurements. European Journal of Applied Physiology, 2007;101(6):743-51. PubMed 17849143
  • Vesterinen V, Nummela A, Heikura I, et al. Individual endurance training prescription with heart rate variability. Medicine & Science in Sports & Exercise, 2016;48(7):1347-54. PubMed 26909534
  • Düking P, Zinner C, Reed JL, Holmberg HC, Sperlich B. Predefined vs data-guided training prescription based on autonomic nervous system variation: a systematic review. Scandinavian Journal of Medicine & Science in Sports, 2020;30(12):2291-304. PubMed 32785959
  • Düking P, Zinner C, Trabelsi K, et al. Monitoring and adapting endurance training on the basis of heart rate variability monitored by wearable technologies: a systematic review with meta-analysis. Journal of Science and Medicine in Sport, 2021;24(11):1180-92. PubMed 34489178
  • Manresa-Rocamora A, Sarabia JM, Javaloyes A, Flatt AA, Moya-Ramón M. Heart rate variability-guided training for enhancing cardiac-vagal modulation, aerobic fitness, and endurance performance: a methodological systematic review with meta-analysis. International Journal of Environmental Research and Public Health, 2021;18(19):10299. PMC8507742
  • Pietilä J, Helander E, Korhonen I, Myllymäki T, Kujala UM, Lindholm H. Acute effect of alcohol intake on cardiovascular autonomic regulation during the first hours of sleep in a large real-world sample of Finnish employees: observational study. JMIR Mental Health, 2018;5(1):e23. PMC5878366
  • Garmin. fēnix 7 Owner's Manual, "Heart Rate Variability Status" (accessed Nov 20, 2026). Garmin
  • Garmin. Understanding the HRV Status on your Garmin smartwatch (accessed Nov 20, 2026). Garmin
  • Apple. HealthKit documentation, heartRateVariabilitySDNN, and Apple Support, "Monitor your heart rate with Apple Watch" (accessed Nov 20, 2026). Developer, Support
  • Polar. Nightly Recharge recovery measurement (accessed Nov 20, 2026). Polar
  • COROS. What is HRV? (accessed Nov 20, 2026). COROS
  • WHOOP. Developer docs, Recovery (accessed Nov 20, 2026). WHOOP
  • Magic Training. Performance page and the product's readiness reading, texts accessed Nov 20, 2026.

Find your plan

Which plan fits your next race?

Answer a few questions about distance, goal and routine and see the plans that fit where you are right now.

Find my plan