Fitbit Charge 6 Accuracy: Top-Tier Heart Rate, Untested Calories

Last updated: October 12, 2026

Fitbit Charge 6 accuracy depends entirely on which number you are looking at. For heart rate, the Charge 6 sits in the top tier of an independent 10-device study, with a median error of 4.5 bpm against a chest-strap ECG, and it held that accuracy in heat and cold. For calories, the only data is a conference abstract with about 20 people per device, where the Charge 6 was off by about 14% on a treadmill test. VO2 max has never been tested on any Fitbit newer than the Charge 2. Steps have one study, in people with lung cancer. Sleep has none. Here is what each study actually found.

If you want to compare the Charge 6 against other devices on heart rate, try the heart rate accuracy tool, then read on for the metrics the tool cannot cover.

I started this expecting a thin evidence file for a three-year-old tracker. The heart rate result surprised me: an inexpensive band from 2023 holding up as well as anything in an independent test, including in a cold chamber where several other devices lost their ability to follow changes in heart rate. The rest of the file is close to empty, and the gap between the two is the real story.

Fitbit Charge 6 accuracy at a glance

Metric	Best independent evidence	Charge 6 result	How much weight it can carry
Heart rate	Gielen 2026, JMIR Formative Research, chest ECG reference	Median error 4.5 bpm (5.5%), top tier of 10 devices	Good. About 10 sessions per device
Calories	Perez 2025, conference abstract, lab gas analysis	About 14% error on a max treadmill test	Low. Not peer reviewed, about 20 people
VO2 max	None for any Fitbit newer than the Charge 2	Not tested	None
Steps	Singh 2026, JMIR Formative Research, video-counted steps	28.5% error, undercounting	Low for most people. 14 adults with lung cancer
Sleep	None against a sleep lab or research device	Not tested	None
HRV, SpO2, skin temperature, stress	None	Not tested	None
ECG (AFib check)	FDA cleared the Fitbit ECG app in 2020	Charge 6 is a supported device	Clearance data predates the Charge 6

Both peer-reviewed studies were independent. Gielen was funded by KU Leuven and Singh by Ohio State, and both declare no conflicts of interest.

Heart rate: where the Charge 6 earns its reputation

Google says the Charge 6 has "the most accurate heart rate on a Fitbit tracker yet." Its launch post claims heart rate during HIIT, spinning and rowing is up to 60% more accurate than before, and the footnote is specific: compared with the Charge 5, based on 90th percentile errors in Google's own 2023 testing, and not applicable to other exercises. That is an internal comparison against the previous model, not an independent test.

The independent test came in 2026. Gielen et al., a KU Leuven study, put 10 wearables through rest, a cognitive stress task, steady walking and stop-start walking in three climate chambers, against heart rate from a Zephyr chest-strap ECG. 45 adults took part.

Tier	Device	Median error (MAPE)	Median error (bpm)	Concordance (CCC)
1	Fitbit Charge 6	5.5%	4.5	0.93
1	Garmin Vivoactive 5	6.3%	5.1	0.83
1	Google Pixel Watch 2	6.7%	4.9	0.87
2	Apple Watch SE	7.3%	5.0	0.70
2	Garmin Vivosmart 5	8.1%	7.0	0.78
3	Polar Ignite 3	11.2%	9.5	0.63
3	Xiaomi Watch 2	11.9%	9.1	0.69
3	Polar Pacer	13.1%	9.7	0.66
4	Oura Ring Gen 3	15.0%	11.0	0.61
4	Fitbit Inspire 3	16.5%	14.3	0.45

We show tiers rather than a numbered ranking for a reason. Each participant wore two watches at a time, so each device was tested 10 times on different people rather than on all 45. Gaps between devices in the same tier are smaller than the chest strap's own error, so the order inside a tier means nothing. What the table does show clearly is the spread: the Charge 6 and the Inspire 3 are both Fitbits, in the same study, at opposite ends. Model matters more than brand.

The Charge 6's average bias was small, 0.7 bpm, but 95% of readings fell within roughly 11 to 13 bpm either side of the reference. That is normal for wrist optical heart rate and is the reason a single reading during a workout should not be over-read.

The cold-weather result most people miss

The climate chambers ran at 23 °C, a hot 36 °C and a cold 10 °C. Error in bpm barely moved for the Charge 6: 4.1, 4.7 and 3.8. The more telling number is concordance, which measures whether the device tracks your heart rate as it rises and falls rather than just landing near the average.

Device	CCC, neutral	CCC, hot	CCC, cold
Fitbit Charge 6	0.94	0.94	0.91
Oura Ring Gen 3	0.78	0.80	0.32
Polar Pacer	0.67	0.78	0.52
Fitbit Inspire 3	0.52	0.58	0.36

Several devices looked fine in the cold on average error, because heart rate was lower and steadier, while their ability to follow changes collapsed. The Charge 6 was the only device that held its concordance in all three conditions. The authors note that climate differences did not reach statistical significance with 10 sessions per device, so treat this as a trend rather than a proven effect. For the wider picture across brands, our post on how accurate your heart rate monitor is covers the full evidence base.

Calories: one abstract and a lot of assumptions

The most searched Charge 6 question is calorie accuracy, and it has the thinnest answer. No peer-reviewed study has tested Charge 6 calories. The only figures come from a 2025 conference abstract from UNLV (Perez and colleagues), which measured energy burn against lab gas analysis during a maximal treadmill test.

Device	Error (MAPE) during a max treadmill test
Samsung Galaxy Watch7	9.42%
Fitbit Charge 6	13.97%
Garmin Forerunner 265S	14.65%
Polar Vantage M3	21.10%
Apple Watch Ultra 2	25.95%

Read this table carefully. The study enrolled 97 adults in total, but each device was worn by roughly 20 different people, so it is not a head-to-head ranking. The figures are unsigned, so the abstract does not say whether the Charge 6 read high or low. It covers one kind of exercise, an all-out treadmill test, and the Charge 6 did not meet the authors' 10% accuracy threshold. Only the Galaxy Watch7 did.

Good heart rate helps a calorie estimate, but it does not validate one. A large share of your daily calorie total comes from your resting metabolism, which the Charge 6 calculates from the height, weight, age and sex in your profile rather than measuring. Everything else that has been published about Fitbit calories comes from older models like the Charge 2 and Charge HR. We cover that lineage, and what it can and cannot tell you, in how accurate Fitbit calorie burn is.

The practical answer is the same one that works for any tracker: calibrate it against your own body. Log what you eat for three to four weeks, track your weight under fixed conditions, and compare the implied burn with what the Charge 6 reported. If the gap is consistent, you have a correction factor. Kygo makes the food side of that quick, with voice, photo, barcode or plain-text logging in 20 to 30 seconds, and pulls your Fitbit data in beside it so the comparison builds itself. Get it on iOS or Android.

VO2 max: untested, and the method just changed

There is no published validation of VO2 max on the Charge 6, or on any Fitbit since the Charge 2. The two Charge 2 studies disagree. Freeberg et al. 2019 (n=30) found the Fitbit score read about 2.9 mL/kg/min high, with a 10.2% error, and a simple non-exercise equation did better at 7.8%. Klepin et al. 2019 (60 people analysed) found acceptable validity in young, fit adults who could run.

Both studies tested an algorithm that no longer exists. When the Fitbit app became Google Health in May 2026, "Cardio Fitness Score" was renamed VO2 max and the method changed. According to Google's current help page:

How Google Health measures VO2 max now	What Google says
When it is measured	"Only during outdoor runs with GPS enabled"
What does not count	Walking, biking and treadmill running
Getting a first reading	One 10-minute outdoor run, or up to three runs in 30 days
How it is calculated	Running pace against heart rate, compared with your resting heart rate

The Charge 6 has built-in GPS and GLONASS, so it can meet that requirement on its own. If you never run outdoors, you will not get a reading. Some owners who do run report the score still not appearing, including one Charge 6 owner in a Google Health community thread about the same problem on the Fitbit Air. We cover that issue in detail in why the Fitbit Air says VO2 max not tracked.

Steps: the first study, in a population you might not expect

The first peer-reviewed Charge 6 step study was published in 2026. Singh et al., at Ohio State, compared the Charge 6 against video-counted steps in 14 adults with lung cancer walking at different speeds in a lab.

Walking speed	Trials	Charge 6 error (MAPE)	Direction
Below 0.6 m/s	14	56.6%	Undercount
0.6 to 0.8 m/s	20	27.1%	Undercount
0.8 to 1.0 m/s	23	23.7%	Undercount
1.0 to 1.2 m/s	31	22.2%	Undercount
1.2 to 1.4 m/s	10	39.1%	Undercount
Above 1.4 m/s	13	28.4%	Undercount

Across a separate set of 126 timed walking trials, overall error was 28.5%.

The Charge 6 undercounted at every speed, and no speed met the 10% benchmark. Short bouts were the hardest: in 16 of 42 five-second walks, it logged no steps at all. This is a small clinical group, often walking slowly, so it does not tell you how the Charge 6 performs for a healthy adult on a normal day. It is still the only Charge 6 step data that exists, and the first time anyone has put a number on this model's step counting at all.

Sleep, SpO2, skin temperature and ECG

Sleep has never been tested on the Charge 6 against a sleep lab or research device. One 2026 study compared it with a questionnaire in breastfeeding mothers, which is not a measure of accuracy.

SpO2 and skin temperature both carry caveats on Google's spec page. SpO2 is "not intended to diagnose or treat any medical condition". The temperature feature shows variation only, is "not intended for medical purposes", and Google notes that big ambient temperature changes may affect it. Neither has been independently tested.

The ECG app was cleared by the FDA in September 2020, from a study of 475 people that found 98.7% sensitivity for AFib and 100% specificity. The FDA summary describes the device only as a Fitbit wrist-worn product, and the Charge 6 did not launch until 2023. The Charge 6 is a supported device for the app, but that clinical data was not collected on it.

Fitbit Charge 6 vs WHOOP

No study has tested the Charge 6 and a WHOOP band together. Other KU Leuven studies tested WHOOP 4.0, but alongside the Charge 5 and Sense 2, so those figures do not transfer. Any accuracy comparison between the two you read online is an inference across separate studies. If you are weighing a screenless tracker, our Fitbit Air vs WHOOP comparison covers sensors and features side by side.

How to get more out of your Charge 6 data

Wear it snug during workouts. Optical heart rate is the Charge 6's strongest metric, and a band that slides around is the easiest way to lose that advantage.

Keep your profile current. Height, weight, age and sex drive the resting part of your calorie total, which is a large share of it.

Run outdoors with GPS on if you want a VO2 max reading at all.

Read weekly trends, not single days. That is true for every metric here, and especially for steps and calories.

Common questions

How accurate is the Fitbit Charge 6?

For heart rate, very good: a median error of 4.5 bpm against a chest-strap ECG in an independent 10-device study, placing it in the top tier. Calories, VO2 max, sleep and most other metrics have little or no independent testing, so their accuracy is unknown.

How accurate is Fitbit Charge 6 calories burned?

There is no peer-reviewed answer. One 2025 conference abstract found about a 14% error during a maximal treadmill test, with about 20 people wearing the Charge 6. Daily totals also depend heavily on your profile data, since resting metabolism is calculated rather than measured.

Is the Fitbit Charge 6 heart rate accurate?

Yes, by current evidence. In Gielen 2026, it had a median error of 5.5% and a concordance of 0.93, in the top tier of 10 devices, and it held that concordance in heat and cold. Individual readings still varied: 95% fell within roughly 11 to 13 bpm of the reference.

Does the Fitbit Charge 6 measure VO2 max accurately?

Nobody has tested it. No Fitbit newer than the Charge 2 has VO2 max validation, and Google changed the method in May 2026. The Charge 6 can now only produce a reading from outdoor runs with GPS.

Is the Fitbit Charge 6 more accurate than WHOOP?

No study has compared them directly, so there is no evidence either way. The studies that tested WHOOP 4.0 used the Charge 5 and Sense 2, not the Charge 6.

Is Fitbit Charge 6 sleep tracking accurate?

It has not been tested against a sleep lab. Treat sleep stages as estimates and focus on trends in total sleep and timing.

Does the Fitbit Charge 6 count steps accurately?

The only study found it undercounted by 28.5% in adults with lung cancer walking in a lab, and missed many very short walks. That group walked slowly, so results for healthy adults may differ.

The bottom line

The Charge 6 is a tracker with one excellent, independently tested metric and a long list of untested ones. Its heart rate holds up with the best devices in the only independent study, and stays reliable in heat and cold. Its calories rest on a conference abstract, its VO2 max on studies of a 2016 tracker using an algorithm that has since been replaced, and its sleep on nothing at all.

That is a reason to trust its heart rate, read its other numbers as trends, and test the calorie figure against your own body before you rely on it.

Ready to see how your food and your Fitbit data actually line up? Kygo connects your meals to your heart rate, sleep and HRV, and shows you the patterns that hold across weeks rather than single days. Get it on iOS or Android.

Key sources

Gielen et al. 2026, JMIR Formative Research, 10 devices, 45 adults, KU Leuven funding, no conflicts declared; Singh et al. 2026, JMIR Formative Research, Charge 6 steps in 14 adults with lung cancer, Ohio State funding, no conflicts declared; Perez et al. 2025, IJES conference abstract, not peer reviewed; Freeberg et al. 2019, mHealth, Charge 2, n=30; Klepin et al. 2019, Med Sci Sports Exerc, Charge 2, 60 analysed; Google's Charge 6 launch post, Charge 6 specs and VO2 max help page, read October 2026; and the FDA 510(k) record for the Fitbit ECG App, K200948.

A note on what this post leaves out

A sleep accuracy figure of "84.6%" for the Charge 6 circulates widely. We could not trace it to any study, so it is not here. You may also see "40% more accurate" heart rate attached to the Charge 6. That claim belongs to the Pixel Watch 2; the Charge 6 claim is 60%, against the Charge 5, in three activities. Finally, we have not used the Gielen results to give the Charge 6 a numbered rank. The study design does not support one, so we report tiers.

---

Disclaimer: Kygo is a personal data aggregation and insights platform designed for informational purposes only. The information provided by Kygo, including correlations, patterns, and trends identified in your data, does not constitute medical advice, diagnosis, or treatment. Always consult a licensed healthcare provider with any questions regarding medical conditions.

---

If you have worn a Charge 6 alongside a chest strap, I would like to know whether its heart rate held up in your workouts the way it did in the lab.