The BMR Calculator estimates your Basal Metabolic Rate — the calories your body burns at complete rest just staying alive — using both of the standard predictive equations, Mifflin-St Jeor and the revised Harris-Benedict, so you can see the two side by side. It then scales your chosen BMR by five activity levels to show maintenance calories (TDEE) from sedentary to extra active. Metric and imperial units are both supported.
The Two Formulas
Mifflin-St Jeor (1990) is the modern clinical default and the formula this calculator leads with:
Harris-Benedict, revised (1984 revision of the 1919 original) uses different coefficients:
TDEE applies the familiar activity multipliers to your (rounded) BMR: sedentary ×1.2, lightly active ×1.375, moderately active ×1.55, very active ×1.725, extra active ×1.9.
Worked Example
A 30-year-old woman, 65 kg, 165 cm:
- Mifflin-St Jeor: 650 + 1,031.25 − 150 − 161 = 1,370.25 → 1,370 calories
- Harris-Benedict: 601.06 + 511.17 − 129.90 + 447.59 = 1,429.92 → 1,430 calories
The formulas disagree by 60 calories — a normal gap. Her TDEE ladder from the Mifflin figure:
| Activity | Multiplier | Maintenance calories | |---|---|---| | Sedentary | 1.2 | 1,644 | | Lightly active | 1.375 | 1,884 | | Moderately active | 1.55 | 2,124 | | Very active | 1.725 | 2,363 | | Extra active | 1.9 | 2,603 |
The gap between formulas also grows with size: for a 40-year-old man at 90 kg and 178 cm, Mifflin gives 1,818 while Harris-Benedict gives 1,921 — 103 calories apart, because the revised Harris coefficients weight body mass more heavily. Neither is "wrong"; both are statistical fits to measured populations, and your true BMR sits wherever your physiology puts it, typically within about ±10% of the Mifflin estimate.
Using the Two Numbers
When the formulas bracket a range, plan against the middle and calibrate with reality: eat at the implied maintenance for two to three weeks and watch the trend of weekly average weights. Stable weight confirms the estimate; steady drift tells you your personal multiplier. This feedback loop is worth more than any refinement of the formula, because BMR is only the first term — the activity multiplier is where most real-world error lives, and it is the term only your own data can pin down.
For calorie targets by goal — deficits and surpluses for loss and gain — the companion calorie calculator builds directly on the Mifflin BMR with goal bands; this page's job is the rate itself and how the two classic equations compare on your numbers.
Limitations
Both equations were fitted on general populations and know nothing about your body composition: lean mass burns more than fat mass at the same weight, so very muscular people are underestimated and people with high body fat are overestimated. If you know your body-fat percentage, a lean-mass formula such as Katch-McArdle (not included here) will usually beat both. Age must be a whole number, and the equations are validated for adults — adolescents and clinical populations need specialist assessment, as do thyroid and other metabolic conditions that move true BMR outside the formulas' error band entirely.
Frequently Asked Questions
Which formula should I use?
Start with Mifflin-St Jeor — validation studies consistently find it the most accurate for the general population, and it is the default here. Use Harris-Benedict as the second opinion it is presented as: if the two are close, confidence rises; if they diverge widely, your body type is at an edge of the fitted populations and real-world calibration matters more than the choice.
Why show TDEE on a BMR calculator?
Because BMR alone is not actionable — nobody eats at BMR, and chronically eating below it is inadvisable. The activity ladder converts the resting figure into the maintenance calories you actually plan around, at every activity level, without a second tool.
Does imperial input change the result?
No. Imperial entries are converted first (pounds × 0.45359237 to kilograms, inches × 2.54 to centimeters) and the same equations run, so a 143 lb, 5 ft 5 in entry reproduces the metric example above apart from rounding in the inputs themselves.