Fitbit step counts are usually accurate within 5 to 10 percent for normal walking, but they can miscount in specific situations—pushing a shopping cart, walking on a treadmill, or moving your arms without walking all confuse the sensor.
Fitbits use a three-axis accelerometer, a small motion sensor that detects movement in all directions. When you walk, the device recognizes the repetitive up-and-down motion of your wrist and converts it into step counts. This works well for steady walking on flat ground, which is why most people find their Fitbit reasonably close to their actual steps.
The problem is that an accelerometer cannot tell the difference between a real step and other arm movements. If you push a shopping cart, swing your arms while standing still, or gesture while talking, your Fitbit may count those as steps. On the flip side, if you walk while holding something—a baby, a coffee cup, a briefcase—your arm stays still, and the device undercounts.
Key Takeaways
- Fitbit step counts are typically within 5 to 10 percent of actual steps during normal walking on flat ground.
- The device miscounts when your wrist does not move with your legs—pushing a cart, carrying something, or walking with your hands in your pockets.
- Treadmill walking often registers as fewer steps because the belt moves beneath you rather than your body moving forward through space.
- Fitbit counts steps based on wrist motion, not foot strikes, so the accuracy depends on how much your arm swings when you move.
Why Fitbits Undercount on Treadmills
Treadmill walking is one of the most common sources of undercounting. When you walk on a treadmill, your body stays in roughly the same place while the belt moves beneath your feet. Your wrist still moves up and down, but the accelerometer is also detecting that your overall position is not changing. Some Fitbit models have learned to account for this, but many still register treadmill steps as 20 to 30 percent lower than they actually are.
If you use a treadmill regularly, you can manually add steps after your workout, or you can log the treadmill session as an exercise in the Fitbit app, which lets you enter the distance or duration separately. The app will then estimate your steps based on your stride length rather than relying on the accelerometer alone.
When Fitbits Overcount Steps
Fitbits are more likely to overcount than undercount in daily life. Any repetitive arm motion can trigger step counts: washing dishes, brushing your teeth, clapping, or even sitting on a bus with bumpy roads can add phantom steps to your total. Some users report gaining 100 to 200 extra steps per day just from normal household movement that is not actually walking.
This is less of a problem if you are using your Fitbit to track trends over time rather than to hit a specific daily target. If you overcount by the same amount every day, your weekly and monthly patterns will still be accurate relative to each other. The issue becomes more noticeable if you are trying to compare your Fitbit count to a pedometer or to steps counted by another device.
How Fitbit Counts Compare to Other Devices
Studies comparing Fitbits to pedometers and smartphone step counters show that Fitbits are generally more accurate than phones (which rely on GPS and accelerometer data in ways that vary by model) but less precise than dedicated pedometers worn at the hip or ankle. A pedometer directly counts foot strikes, while a Fitbit infers steps from wrist motion, so there will always be a gap.
If you wear your Fitbit loosely or on your non-dominant wrist, accuracy drops further. The device needs to move with your arm to detect the rhythm of your walk. A loose band or a watch worn on the wrist that does not swing as much during your gait will undercount.
Factors That Affect Fitbit Accuracy
Your walking style matters more than you might think. People with shorter strides, slower gaits, or who do not swing their arms much while walking will see lower step counts. Older adults and people with certain mobility conditions often find their Fitbits undercount because their arm swing is less pronounced. Conversely, people who swing their arms energetically or who have a longer stride will see counts closer to reality.
The surface you walk on also plays a role. Walking on soft ground, sand, or carpet produces less consistent accelerometer readings than walking on hard floors or pavement. Inclines and declines can also throw off the sensor, since the up-and-down motion of climbing stairs or walking downhill is different from level walking.
How to Check Your Fitbit's Accuracy
The simplest way to test your device is to count your steps manually over a short distance—say, 100 steps—and compare that to what your Fitbit records. Walk at your normal pace, count each step out loud or on your fingers, and then check the Fitbit app. If the numbers are within 5 to 10 steps, your device is performing as expected.
For a longer test, walk a measured distance (a track, a route you know, or a distance measured by a mapping app) and compare the steps your Fitbit counted to the distance you covered. Divide the distance by the number of steps to find your average stride length, then use that to estimate whether the count makes sense. If your Fitbit says you took 1,000 steps but you only walked a quarter mile, something is off.
When to Manually Log or Adjust Your Steps
If you know your Fitbit is systematically undercounting or overcounting, you have options. In the Fitbit app, you can manually log exercise sessions and enter the distance or duration, which overrides the step count for that period. You can also adjust your stride length in the device settings—a longer stride length will convert the same number of accelerometer pulses into fewer steps, and vice versa.
For activities where you know the Fitbit will struggle—pushing a stroller, carrying groceries, or working out on a treadmill—logging the activity separately is more accurate than relying on the step count alone. The Fitbit app has categories for dozens of activities, and you can enter the duration and intensity, which gives you credit for the work without depending on the accelerometer.
Frequently Asked Questions
Why does my Fitbit count steps when I am not walking?
The accelerometer detects any repetitive arm motion, not just walking. Washing dishes, clapping, bumpy car rides, and even fidgeting can register as steps. This is normal and happens with most wrist-worn devices. If the phantom steps are significant, check that your band is snug enough to move with your wrist.
Is my Fitbit broken if it counts fewer steps on the treadmill?
No. Treadmill undercounting is a known limitation of wrist-worn accelerometers. Your device is working correctly; it is just not designed to account for the fact that you are staying in one place. Log treadmill sessions as exercise instead of relying on step count.
Can I trust my Fitbit to compare my activity to someone else's?
Not exactly. Since accuracy depends on arm swing, stride length, and walking style, two people walking the same route may get different step counts. Fitbits are better for tracking your own progress over time than for comparing counts between different people.
Does wearing my Fitbit on my non-dominant wrist change the accuracy?
Yes, usually for the worse. Your non-dominant arm typically swings less during walking, so the accelerometer detects less motion. If you want the most accurate count, wear your Fitbit on your dominant wrist, or adjust your stride length setting to compensate.
What is the best way to get an accurate step count from my Fitbit?
Wear it snugly on your dominant wrist, walk at a steady pace on flat ground, and let your arms swing naturally. For activities where the device struggles—treadmills, pushing carts, or carrying things—log them as exercise in the app instead of relying on the step counter.