Why Fitness Trackers Keep Overstating Your Calorie Burn

The Invisible Math Error on Your Wrist

Sarah had been doing everything right. Six days a week at the gym, a steady calorie deficit tracked down to the last almond — and after two months, she'd lost exactly nothing. Her Fitbit said she was burning 2,800 calories on workout days. Her food log was pristine. The math, at least on paper, pointed to a weekly deficit of nearly 3,500 calories. That's a pound a week. Where was it going?

Her trainer had a blunt answer: nowhere, because the deficit was never real. "I see this constantly," says personal trainer Marcus Webb, who works with clients in Chicago. "People eat back their 'earned' calories based on what the watch says, and the watch is lying to them. Not because it's a bad device — it's just not designed to be a metabolic chamber."

This isn't a fringe complaint. A growing body of independent research suggests that consumer fitness wearables routinely overestimate calorie expenditure during exercise — sometimes by a margin that would make your cardiologist wince.

What the Studies Actually Found

The most-cited investigation on this topic came out of Stanford University in 2017, led by researchers Euan Ashley and Anna Shcherbina. They strapped seven popular devices — including the Apple Watch, Fitbit Surge, and Basis Peak — onto 60 volunteers and compared the readouts against indirect calorimetry, the gold-standard lab method for measuring energy expenditure. The results were uncomfortable reading for the wearable industry.

Heart rate error ranged from about 5 percent for the best performer to nearly 35 percent for the worst. But calorie error was far worse. Not a single device measured energy expenditure within an acceptable margin of error. The least inaccurate device was off by 27 percent. The worst missed by 93 percent. That's not a rounding error — that's a different universe of measurement.

A 2020 study published in the Journal of Sports Sciences tested devices during common activities like running, cycling, and resistance training. Researchers at Aberystwyth University found that overestimation was particularly pronounced during strength work, where wrist-based optical heart rate sensors struggled to accurately capture the non-linear cardiovascular demands of lifting. During a 30-minute weight session, some devices reported calorie burns 40 to 50 percent higher than actual measured output.

More recently, a 2022 analysis from the University of Colorado examined whether the newer generation of smartwatches — with their additional sensors and updated algorithms — had narrowed the accuracy gap. The conclusion was cautiously optimistic but ultimately deflating: accuracy had improved for walking and outdoor running, but the fundamental problem of overestimating remained, especially during high-intensity interval training and any activity that involves significant arm movement that mimics the motion of walking.

How the Devices Build Their Estimates

Understanding the error means understanding the method. Most wearables use optical photoplethysmography — a green LED that bounces light off your skin to detect blood volume changes in your capillaries, which they correlate to heart rate. From heart rate, they estimate oxygen consumption. From estimated oxygen consumption, they derive calories burned. Each step in that chain introduces potential error, and those errors compound.

Device manufacturers then layer in personal data — your age, weight, height, and sometimes VO2 max estimates — to refine the output using proprietary algorithms. These algorithms are trained on datasets that tend to skew toward certain demographics, which is part of why accuracy can vary significantly by body type, skin tone, and fitness level.

The fundamental problem is that calorie burn is not a simple function of heart rate. Two people with identical heart rates during a workout can have meaningfully different metabolic costs, depending on muscle mass, movement efficiency, and individual physiological variation. A wristwatch cannot measure any of this directly.

What Manufacturers Say

Ask a wearable company about accuracy and the response tends to be diplomatically evasive. Fitbit, Garmin, and Apple have all pointed to internal validation studies showing strong correlation between their devices and lab-grade equipment — but the methodology and populations used in those studies are rarely made fully public.

Apple's health documentation describes calorie figures as "estimates" and notes that "the accuracy of calorie and activity data depends on your personal information and movement data." Garmin similarly frames its burn figures as estimates based on physiological models. These are not admissions of failure — they're accurate descriptions of what the technology actually does. The problem is that the product design — large, bold calorie numbers presented in real time — communicates certainty that the underlying science cannot support.

"Marketing and reality are doing two different jobs," says Dr. Lorraine Pemberton, an exercise physiologist who has consulted for sports technology companies. "The number has to feel authoritative, or the app loses its appeal. But authoritative and accurate are not the same thing."

The Specific Problem for Dieters

For someone using a tracker primarily to count steps or monitor sleep, the calorie error is mostly cosmetic. But for someone in a calorie deficit trying to lose weight, the overestimation creates a specific and insidious trap.

Nutrition apps like MyFitnessPal integrate directly with wearables and automatically add "active calories" back to your daily allowance. If you burned 400 calories on a run, the app adjusts upward — and most people eat some or all of those calories back. If the device was actually overstating by 30 percent, you burned 280 calories, not 400. That 120-calorie phantom gap, repeated across a week, erases roughly half a pound of would-be weight loss.

High-intensity intervals: Devices consistently overcredit HIIT, partly because post-exercise oxygen consumption isn't captured after the session ends.
Strength training: Wrist motion and elevated heart rate both contribute to inflated readings during weight sessions.
Spinning and cycling: Lower arm movement means heart rate capture is cleaner, but cycling efficiency varies so widely between individuals that calorie models struggle.
Swimming: Most optical sensors perform poorly when wet; waterproof devices often use stroke counts and pre-set MET values that don't account for individual stroke efficiency.

Walking, oddly, is where most modern devices perform best. The motion is regular and well-understood, and step-counting algorithms have been refined over a decade. Even here, though, gradient, load, and individual gait can introduce meaningful variation.

Recalibrating Your Relationship With the Number

None of this means you should bin your tracker. Wearables are genuinely useful for tracking trends, building movement habits, monitoring resting heart rate over time, and keeping a rough sense of daily activity. The research problem is not with the devices as habit tools — it's with treating their calorie output as a precise measurement to be entered into a weight-loss spreadsheet.

Here are the adjustments that exercise physiologists and registered dietitians most commonly recommend:

Apply a personal correction factor. If you've been eating at what looks like a 500-calorie deficit and seeing no movement on the scale over four weeks, your actual deficit is probably closer to zero. Work backward from results, not from device output.
Don't eat back all exercise calories. A conservative approach is to eat back no more than 50 percent of what the device claims. This buffers against overestimation while still accounting for genuine increased energy needs.
Use heart rate data, not calorie data. Zone distribution — how much time you spent in fat-burning versus cardio versus peak zones — is far more reliably measured than total calorie burn. Use it to guide workout intensity, not to justify extra servings.
Consider a chest strap for serious sessions. Devices like Polar H10 or Garmin HRM-Pro use electrical signal capture rather than optical sensors, which produces substantially more accurate heart rate data — and therefore better calorie estimates — during intense exercise.
Trust body composition trends over short windows. Weight, measurements, and how clothes fit over four to six week periods tell you more than any single device reading.

The Bigger Picture

The fitness tracker market crossed $60 billion globally in 2023 and shows no sign of slowing. Devices are getting more sophisticated — continuous glucose monitors are now entering the consumer space, and some watches claim to estimate blood oxygen, stress, and even hydration. The temptation to trust these numbers grows as the interfaces become more polished and the marketing becomes more confident.

But the lesson of a decade of independent research on wearable accuracy is consistent: the gap between what these devices appear to measure and what they actually measure remains wide, especially when it comes to calorie expenditure. That doesn't make them useless. It makes them useful in a different way than most people assume.

Your tracker is best understood as a motivational and behavioral tool that offers directional physiological signals. The moment you start making precise nutritional decisions based on its calorie readout, you're asking it to do something it was never designed to do reliably — and your waistline may end up paying for the confusion.

Sarah, for what it's worth, eventually stopped eating back her exercise calories. Within six weeks, she'd dropped four pounds. The math, once corrected for what her wrist had been making up, finally worked.