September 21, 2026, 2:03 am | Read time: 6 minutes
Many people rely on the calorie counts from their smartwatches for weight loss or training. However, a new study from the U.S. with popular smartwatch models suggests that the displayed calorie consumption is often significantly inaccurate. Particularly problematic: As body fat percentage increases, the estimates become less accurate. How the study was conducted and how you can still use your smartwatch effectively.
Why were the smartwatches examined?
Smartwatches and fitness trackers have become part of the daily health and training routine for many people. They record heart rate, physical activity, and estimated calorie consumption, among other things. Many users like to use the energy consumption data for weight loss or training planning. However, the accuracy of these estimates has been controversial for years. Previous studies have already pointed to sometimes significant errors.1,2 However, it was unclear which factors influence these inaccuracies.
The current study therefore specifically examined the role of body fat percentage and skin tone. The background is that the optical sensors of many smartwatches may measure differently depending on skin pigmentation. The aim of the study was to compare the calorie consumption estimates of four widely used smartwatch models with an established laboratory method for measuring actual energy consumption.3
4 Smartwatches in the Test
The study involved 58 adults aged 18 to 50. All participants were of Hispanic background and belonged to skin types III to V.
To examine calorie consumption, participants completed a standardized workout on a recumbent bike, alternating between moderate and intense phases. During the workout, actual energy consumption was recorded with a COSMED K5 metabolic analyzer. This method is based on indirect calorimetry and is considered the gold standard for determining energy consumption. At the same time, participants wore four different smartwatches:
- Apple Watch Series 8
- Fitbit Sense 2
- Samsung Galaxy Watch 5
- Garmin Forerunner 955
Researchers then compared the calorie values calculated by the watches with the lab values. Additionally, body fat percentage, body measurements, and skin type were recorded. Using statistical models, the authors analyzed which factors were associated with measurement errors.
Fitness Watches Fail–One is Completely Unusable
The results show that all the smartwatches examined misestimated calorie consumption, sometimes significantly. The average deviation from the lab values was:
- Apple Watch: 21.6 kcal
- Garmin Forerunner: 68.6 kcal
- Samsung Galaxy Watch: 56.8 kcal
- Fitbit Sense 2: 3.1 kcal (but high error rate)
The seemingly good Fitbit result should be interpreted with caution. The device was noted for numerous mismeasurements and in some training sessions provided no calorie values at all, reported only one calorie burned for an entire workout, or overestimated actual energy consumption by more than 450 percent. Several of these measurements were therefore excluded from the analysis.
Thus, the actual winner is the Apple Watch with the smallest deviations. Garmin and Samsung significantly overestimated energy consumption. The average percentage errors often ranged between about 15 and 25 percent, depending on the device. According to the authors, this means that a displayed burn of 500 calories can actually be much lower, around 350 calories.
Also interesting: Don’t be afraid of strength training! Here’s how to get started easily
How Did Body Fat Percentage and Skin Type Affect the Result?
The most important influencing factor was body fat percentage. As body fat increased, measurement errors significantly rose for all devices. This correlation was particularly pronounced with Fitbit and Garmin. For the skin types examined, however, researchers found no connection with the accuracy of calorie measurement.
According to study leader Jason Kostrna, the team does not yet have a definitive answer as to why the trackers have difficulties, especially with higher body fat percentages. He suspects that the algorithms used to estimate calorie consumption may play a role. “When companies develop these algorithms, we don’t know where the test data comes from,” he said in a University of Florida press release.4
Nutritional Recommendations for Strength Athletes – What and When You Should Eat to Build Muscle
Study Shows Link Between What Fathers Eat and Baby’s Birth Weight
How to Use Your Smartwatch Effectively Anyway
The results are especially important for people who use smartwatch data for weight control. Those who want to lose weight but eat more due to distorted data could be thwarted by the fitness watch. Kostrna sums it up: “You can’t consider the displayed calorie consumption number as an exact value, because it isn’t.” In his view, such errors can add up to several hundred calories over weeks.
For users, this doesn’t mean that smartwatches are worthless. The researchers emphasize that the calorie data should be understood as a rough guide and not as an exact measurement. It’s advisable to include a safety buffer in the values, especially if you have a high body fat percentage. At the same time, the authors warn against eating too little in response. The body still needs enough energy for recovery and regular training.
According to the researchers, other measurements from many smartwatches could be more useful than calorie consumption. These include heart rate, time in specific heart rate zones, or long-term developments over weeks and months. Kostrna therefore recommends: “Observe your own trends over time.” Those who want to track their fitness development should focus more on long-term trends than on the calorie count of a single workout.
Study Assessment and Outlook
The study is one of the more comprehensive investigations that directly compares several well-known smartwatch brands with a gold standard of metabolic measurement. Strengths include the controlled lab conditions, standardized exertion, and consideration of body fat percentage as a possible influencing factor.
However, there are some limitations. Participants only completed a workout on a recumbent bike. Whether the results are transferable to other activities remains open. Additionally, only Hispanic adults were examined, and the number of participants with skin type V was small.
Since manufacturers do not disclose their algorithms, it is also not possible to determine which technical factors cause the measurement errors. The authors suspect that the models may not be sufficiently tailored to individuals with higher body fat percentages. Study leader Kostrna is currently investigating how accurately smartwatches estimate calorie consumption during classic strength training sessions.
Regardless, the study makes it clear that calorie data from current smartwatches should be taken with caution. For training planning, long-term developments, such as heart data, can be more meaningful than the displayed calories of a workout.