From left, Gabriel Cabeda Ramirez, a bachelor’s student in sport and exercise science; teaching and learning kinesiology Ph.D. student Daimien Campos; and FIU researcher Jason Kostrna, associate professor of kinesiology and exercise science, participate in a demonstration of the study protocol on Friday, Aug. 21, 2026, at Florida International University in Miami. Cabeda Ramirez and Kostrna reviewed readings from four smartwatches worn by Campos as he rode a stationary bike. The study evaluated how accurately popular smartwatches estimate calories burned during exercise.
Taimy Alvarez/Florida International University
Taimy Alvarez/Florida International University
Don’t trust your smartwatch’s calorie burn count. It’s likely lying to you.
That’s what Florida International University researchers learned while testing popular smartwatches from Apple, Fitbit, Samsung and Garmin as part of a new study analyzing how effective the wearable tech is for tracking calories.
Turns out, the watches don’t do a very good job and tend to significantly overestimate how many calories people burn during exercise, according to the study. And the more body fat you have, the less accurate your calorie burn is, according to Jason Kostrna, the study’s lead author and an associate professor of kinesiology and exercise science at FIU.
“America does have an obesity epidemic, and a portion of that weight management strategy is physical activity,” Kostrna told the Miami Herald in a phone interview Thursday. “The values that we get from our watches when we do exercise oftentimes informs what we choose to eat. They inform what we do for the rest of the day.”
Seeing a high calorie burn count might make you feel good initially. But because the watches are often providing inaccurate calorie estimations, people are unable to “make accurate decisions about their health and physical activity and nutrition,” Kostrna said.
FIU smartwatch calorie tracking study: The results
The study put four popular smartwatches — Apple Watch Series 8, Fitbit Sense 2, Samsung Galaxy Watch 5 and Garmin Forerunner 955 — to the test against a COSMED K5 metabolic analyzer (think: a bulky face mask with wires), the “gold standard” in clinical labs to estimate calories burned.
Participants — a group of 58 people (31 women, 27 men) ranging in age from 18 to 50 — wore all four watches and the COSMED metabolic analyzer simultaneously while exercising on a stationary bike at an FIU lab. Each session was 60 minutes, with the actual exercise portion of the session done in about 20 minutes, with participants cycling hard and fast enough to get their heart rate into the moderate and vigorous zone.
The findings: “All of them struggled,” Kostrna said, noting that the smartwatches’ errors were often in the 15% to 25% range.
FIU researcher Jason Kostrna, associate professor of kinesiology and exercise science, wears the four smartwatches tested in his study — the Apple Watch Series 8, Fitbit Sense 2, Samsung Galaxy Watch 5 and Garmin Forerunner 955 — during a portrait session Friday, Aug. 21, 2026, at Florida International University in Miami. Kostrna led a study that found the devices overestimated calories burned during exercise, with errors increasing among participants with higher body-fat percentages. Taimy Alvarez/Florida International University Taimy Alvarez/Florida International University
Out of the wearables tested, Apple “technically” was the most accurate, he said. Garmin and Samsung overestimated the most. Fitbit was a wild card, with the watch sometimes failing to count any calories, and at other times giving what seemed like implausible and faulty results, including burning just one calorie in an entire workout, according to an FIU news release on the study. The research was published this summer in PLOS One, a peer-reviewed multidisciplinary journal.
For the person wearing the watch, that means the wearable might estimate you burned 500 calories during a workout, but in reality, you actually burned around 350 calories, according to the news release.
“If you multiply that out over a week, that’s a major concern,” Kostrna said, noting it completely skews your knowledge of how many calories you’ve actually burned, potentially impacting the food or exercise decisions you make on your weight-loss journey.
“What our study really found is that the body-fat percentage is so important in the accuracy of these devices,” Kostrna said.
Body-fat percentage determines how much of a person’s weight is due to actual fat, a measure that people can get at some gyms. It’s different from body mass index (BMI), which uses height and weight to estimate a person’s body fat. A football player, for example, might have the same height and weight as someone who doesn’t exercise, but their body-fat percentage will be vastly different, he said.
A person with 15% body fat, typically considered healthy for men and women, will get a fairly accurate calorie burn estimate from their smartwatch, he said. On the other hand, a person with 30% to 35% body fat has a higher chance of receiving an overestimated calorie burn count.
That could lead to overeating and demotivation when the scale doesn’t move as much as you’d expect.
The goal of the study
Kostrna, who has worked as a personal trainer and also exercises himself, noticed that smartwatches’ calorie estimates “can vary wildly,” even if a person exercises in similar conditions, with a similar heart rate. He’s had clients who, at the end of the workout, looked down at their watch to see they burned one calorie, which he knew wasn’t true. Kostrna, who has used various smartwatches and currently uses one from Garmin, noticed large discrepancies in his own exercises, too.
And as the popularity of wearable devices increase, and the tech improves, more doctors and patients are using the tech to help track and monitor heart rate and other health data. Kostrna wanted to figure out how effective the watches actually were, especially for those who were trying to lose weight.
Previous research had indicated the smartwatches could experience errors, such as in people with darker skin tones, he said. He also knew that while the tech’s sensors are effective in measuring heart rate and steps, the wearables currently don’t have a “direct way” of measuring calorie burn.
And obesity rates are continuing to rise in adults and children. A new report published earlier this year in the American Heart Association’s journal Circulation found that about 40% of adults and 20% of people under the age 20 in 2023 met the criteria for obesity, a chronic condition that also increases the person’s risk for other conditions, including heart disease, stroke and Type 2 diabetes.
The limitations
To get the most accurate reading possible, Kostrna used BMI to help screen initial participants and then used a more accurate body-fat percentage exam for their individual sessions. He said all participants were Hispanic to examine how accurate the watches are for South Florida’s Hispanic-majority community. Some of the participants were FIU students, and none had forearm tattoos (tats, it turns out, can interfere with smartwatch sensors).
But the study did have some limitations. That included participants wearing multiple devices simultaneously, which could have potentially affected sensor accuracy, and only having each participant do one exercise session, a decision that restricted researchers’ ability to see if the wearables would learn and adapt as the user did more exercises, according to the study. It’s also worth noting that newer watches have been released since the study was completed, so it’s possible the calorie tracking software has improved.
It’s not clear why smartwatches tend to overestimate calorie burning for people with higher body-fat percentages. Kostrna thinks it must be algorithm-related since the watches are unable to directly measure energy expenditure directly. He now plans to test how accurate the watches measure calorie burning for strength-training exercises. The Herald has contacted Apple, Samsung, Google (which now owns Fitbit) and Garmin for comment on the study’s findings.
Tips to use your fitness tracker effectively
To be clear, this doesn’t mean you can’t rely on your wearables at all. Kostrna acknowledges that the tech continues to improve and is considered to be generally accurate for heart rate, step count and GPS distance. But the study results overall indicate that wearables shouldn’t be the only way to track your calorie deficit or calorie surplus goal.
Instead, Kostrna recommends using the smartwatches’ calorie tracker as a rough guide and to round down how many calories the watch claims you burned because there’s a chance the wearable has a “significant error” in its estimate.
For example, if his watch says he burned 500 calories in a workout, “I will say that number could be anywhere from like 300 to 600,” he said.
He also recommends tracking the intensity of your workouts and to “be aware of how your body feels post-workout as well.”
“Don’t let the data override how you feel about your body,” he said. “It can be an extra data point but don’t make it be the only data point.”
Miami Herald
Michelle Marchante covers the pulse of healthcare in South Florida and also the City of Coral Gables. Before that, she covered the COVID-19 pandemic, hurricanes, crime, education, entertainment and other topics in South Florida for the Herald as a breaking news reporter. She recently won first place in the health reporting category in the 2025 Sunshine State Awards for her coverage of Steward Health’s bankruptcy. An investigative series about the abrupt closure of a Miami heart transplant program led Michelle and her colleagues to be recognized as finalists in two 2024 Florida Sunshine State Award categories. She also won second place in the 73rd annual Green Eyeshade Awards for her consumer-focused healthcare stories and was part of the team of reporters who won a 2022 Pulitzer Prize for the Miami Herald’s breaking news coverage of the Surfside building collapse. Michelle graduated with honors from Florida International University and was a 2025 National Press Foundation Covering Workplace Mental Health fellow and a 2020-2021 Poynter-Koch Media & Journalism fellow.
Support my work with a digital subscription