The problem: steps do not directly map to calories
A fitness tracker tells you that you walked 10,000 steps and burned 347 calories. Only one of those numbers is a measurement. The step count came from an accelerometer. The 347 calories is a chain of three estimates stacked on top of that measurement — and each estimate has a real error band.
The chain is: steps → distance → time → calories. Each arrow is an assumption. Distance requires a stride length, which the tracker estimates from your height (or a default). Time requires a speed, which the tracker assumes based on whether it thinks you are walking or running. Calories require an intensity factor, which the tracker looks up from a table based on the activity it inferred. None of these are measured. They are all inferred.
The result is a number that is useful for tracking relative change — did you burn more this week than last — but unreliable as an absolute truth. You did not burn exactly 347 kcal. You burned something in the neighborhood of 347, with a band of plus or minus 20 percent depending on your actual pace, your fitness level, the terrain, and how your stride differs from the population average.
The Step Count to Calories Calculator makes this chain explicit. It uses the standard MET formula, estimates stride from your height, assumes a fixed speed for walking and running, and shows you the distance, time, and calories that fall out. It does not pretend the number is exact. It is a model, and like every model, it is wrong in a specific, useful way.
Fastest path
Open the Step Count to Calories Calculator. Toggle metric or imperial. Enter your step count, weight, and height. Pick walking or running. The tool shows four numbers: calories burned, distance (km and miles), and active time. Switch to the Food Equivalents tab to see what the calorie number corresponds to in familiar foods. Switch to the Step Goal Calculator tab to reverse the math — enter a calorie target and the tool tells you how many steps you would need.
MET: the unit under the formula
MET stands for Metabolic Equivalent of Task. One MET is your resting metabolic rate — the energy you burn sitting still. By convention, 1 MET is 3.5 milliliters of oxygen per kilogram of body weight per minute, which works out to roughly 1 kilocalorie per kilogram per hour. A 70 kg person burns about 70 kcal per hour doing nothing. That is the baseline.
Activities are rated in METs as multiples of that baseline. Walking at 5 km/h is about 3.5 MET — you burn 3.5 times your resting rate. Running at 9 km/h is about 8.0 MET. Cycling at a moderate pace is about 6.0 MET. Sleeping is about 0.95 MET. The Compendium of Physical Activities, maintained since 1993 and updated periodically, catalogs roughly 800 activities with their MET values.
The calorie formula is direct:
calories = MET × weight(kg) × time(hours)
A 70 kg person walking for one hour at 3.5 MET burns 3.5 × 70 × 1 = 245 kcal. The same person running for one hour at 8.0 MET burns 8.0 × 70 × 1 = 560 kcal. The formula is linear in weight (a heavier person moves more mass, so burns more) and linear in time (longer activity, more energy).
The thing to notice is that the formula needs time, not steps. A step count alone tells you nothing about energy — you could have taken those steps in 30 minutes of brisk walking or 90 minutes of strolling. The intensity is what drives the calorie burn, and the tool has to infer the intensity from the activity type and the assumed speed.
How the tool derives time from steps
The tool never asks you how long you walked. It derives the time from the step count, your height, and an assumed speed.
First, it estimates your stride length from your height. The walking stride factor is 0.413 — your stride in meters is approximately 0.413 × your height in centimeters / 100. A 170 cm person has a walking stride of about 0.70 meters. For running, the factor is 0.54, giving the same person a running stride of about 0.92 meters.
The 0.413 coefficient is a pedometer approximation from the 1990s, derived by averaging measured stride lengths across a sample of walkers. It is a population average, and individuals vary by 10 to 15 percent around it. Your actual stride depends on your leg length, your cadence, your pace, and whether you are walking on flat ground or uphill. The tool cannot see any of that, so it uses the average.
Second, it computes distance: steps × stride. Ten thousand steps at 0.70 meters is 7,000 meters, or 7 km.
Third, it divides distance by an assumed speed to get time. Walking is assumed at 5 km/h, running at 9 km/h. Seven kilometers at 5 km/h is 1.4 hours, or 84 minutes.
Finally, it applies the MET formula: 3.5 × 70 × 1.4 = 343 kcal. That is the number you see.
The assumed speed is the weakest link in the chain. A slow walker at 3 km/h and a brisk walker at 6 km/h both have the tool report the same per-step calorie credit, because the tool cannot tell them apart. The brisk walker is actually burning more per minute (because they are moving faster, which means higher intensity, which means higher MET), but the tool uses the same 3.5 MET for both, and the faster walker gets a shorter time, which cancels the speed difference out. The net effect is that the tool's estimate is reasonable for a 5 km/h walker and wrong for anyone much faster or slower.
Walking vs running: why the METs differ
Walking at 5 km/h is 3.5 MET. Running at 9 km/h is 8.0 MET. You might expect the MET to scale with speed, but running is less efficient than walking — it has more vertical oscillation (you bounce off each step), more ground-contact force, and more arm drive. At the same speed, running burns more calories than walking, which is why the MET table has a discontinuity at the walk-run transition (around 7-8 km/h for most people).
The tool captures this with two hard-coded pairs: (MET 3.5, speed 5 km/h) for walking and (MET 8.0, speed 9 km/h) for running. It does not let you pick a speed. If you run at 10 km/h, the tool underestimates your burn (it assumes 9). If you walk at 6 km/h, the tool overestimates (it assumes 5, so it gives you more time at the same MET). The errors are in the 10-20 percent range for moderate deviations from the assumed speeds.
Gross vs net: the baseline you are already burning
The MET values used here are gross — they include the resting baseline. When the tool says you burned 343 kcal walking, that is the total energy you spent during the walk, including the energy you would have spent anyway just sitting still.
If you want the net energy — the activity-only calories, the surplus above resting — you subtract the resting component. The resting rate is 1 MET, so the net MET for walking is 3.5 - 1 = 2.5, and the net calories are 2.5 × 70 × 1.4 = 245 kcal. That is the energy the walk actually cost you beyond what you would have burned anyway.
The distinction matters for two reasons. First, if you are comparing the calorie burn of an activity to the calories in a meal, the gross number overstates the contribution — you would have burned 98 kcal of that 343 just by sitting on the couch for the same 84 minutes. Second, if you are using the number to plan a deficit for weight loss, the net is the honest figure. The tool reports gross, which is what most fitness trackers also report, but it does not label it as such.
Why the number is always rough
Step-based calorie estimates have a real error band, and it comes from several places.
Stride length varies. The 0.413 × height formula is a population average. Your actual stride is shorter when you are tired, longer when you are walking downhill, shorter on stairs. A 10 percent stride error becomes a 10 percent distance error and a 10 percent calorie error.
Speed is assumed, not measured. A slow walker and a brisk walker get the same per-step credit, which is wrong in opposite directions. A fitness watch with GPS measures actual speed and does better.
Heart rate tracks effort, steps do not. The thing that actually drives calorie burn is oxygen consumption, which tracks heart rate. A heart-rate-based estimate (the kind a fitness watch with an optical sensor produces) is usually within 5-10 percent for most people. A step-based estimate is usually within 20-30 percent. If you care about the number, the watch wins.
Fitness level changes efficiency. A trained runner burns fewer calories per kilometer than an untrained runner at the same speed, because their movement is more efficient. The MET table does not know your training status.
Terrain and incline are not modeled. Walking uphill burns 30-50 percent more than walking on the flat; walking downhill burns slightly less. The tool assumes flat ground. A real hike with elevation changes can be off by 50 percent or more.
Age, sex, and body composition are not in the formula. The tool collects age and gender but does not use them — the MET formula uses only weight. Real energy expenditure varies with lean body mass (muscle burns more than fat), age (metabolic rate declines about 1-2 percent per decade after 30), and sex (men typically have more lean mass than women at the same weight). The tool's number ignores all of this.
The honest summary: the tool's number is a model output, not a measurement. It is useful for comparing days (did I move more this week than last) and for rough planning (is this walk enough to offset that meal). It is not useful for exact accounting, and it should not be used to justify eating a specific quantity of food.
Gotchas
- Age and gender are collected but not used. The tool asks for both, but the formula uses only weight, height, steps, and activity type. This can mislead you into thinking those factors are accounted for. They are not. If your actual metabolic rate differs from the population average for your age and sex, the tool's number is off in the direction of that difference.
- The speed is assumed, not measured. Walking is fixed at 5 km/h and running at 9 km/h. If you walk at 6 km/h, the tool overestimates your time and your calories. If you run at 11 km/h, the tool underestimates both. A fitness watch with GPS measures your actual speed and does better.
- The number is gross, not net. It includes the resting baseline you would have burned anyway. If you are comparing to a meal or planning a deficit, subtract about 1 kcal/kg/hr (your resting rate) from the result to get the activity-only figure.
- Stride is estimated, not measured. The 0.413 × height formula is a population average with a 10-15 percent individual variation. If you have unusually long or short legs for your height, or if your natural cadence is unusual, the estimate is off by that amount.
- Terrain and incline are not modeled. Uphill burns 30-50 percent more; downhill burns slightly less. The tool assumes flat ground. For a real hike, expect the tool to be off by the elevation effect.
- The accuracy band is roughly plus or minus 20-30 percent. Step-based calorie estimates are rougher than heart-rate-based estimates. Use the number for relative tracking (more this week than last), not for exact accounting. If you need a precise number, a fitness watch with a heart-rate sensor is more accurate than a step count.
Summary
- Steps do not directly map to calories. The chain is steps → distance → time → calories, and each arrow is an estimate with a real error band. The step count is the only thing measured; everything else is inferred.
- MET (Metabolic Equivalent of Task) is the intensity unit. 1 MET is resting, about 1 kcal/kg/hr. Walking is 3.5 MET, running is 8.0 MET. The formula is calories = MET × weight(kg) × time(hours). The tool derives time from steps × stride / assumed speed, not from a measured duration.
- Stride is estimated as 0.413 × height for walking and 0.54 × height for running — a 1990s pedometer average that varies 10-15 percent between individuals. Speed is assumed at 5 km/h for walking and 9 km/h for running, regardless of your actual pace.
- The tool reports gross calories (including the resting baseline), not net (activity-only). For weight-loss planning, subtract about 1 kcal/kg/hr to get the net figure. Walking and running have different METs because running is less efficient — more vertical oscillation and ground-contact force at the same speed.
- The number is always rough: stride varies, speed is assumed, heart rate (which tracks actual effort) is not measured, fitness level changes efficiency, terrain is not modeled, and age and sex are collected but not used. Expect a plus-or-minus 20-30 percent error band. Use the number for relative tracking, not exact accounting. A fitness watch with a heart-rate sensor is more accurate than a step count.
- Use the Step Count to Calories Calculator for the estimate, the Heart Rate Zone Calculator for the effort-based companion metric, the BMI Calculator for the body-composition context, and the Daily Water Intake Calculator for the matching hydration figure.