BMR
The energy you would spend lying still all day. Three equations, side by side, because they disagree — and the honest answer is a range, not a figure.
1,700kcal
—
- Mifflin–St Jeor (1990)
- —
- Harris–Benedict (1984)
- —
- Katch–McArdle
- —
- Cunningham (1980)
- —
- Spread between them
- —
- Maintenance at this activity
- —
- BMR as a share of that
Enter a body fat percentage to add the two equations that work from lean mass. Without one, they cannot be calculated.
The headline is Mifflin–St Jeor. The band beside it is ±10%, which is roughly how far an individual sits from the population regression even when the equation is the right one. Measured BMR needs indirect calorimetry, not arithmetic.
What basal metabolic rate measures
BMR is the energy your body spends keeping itself running with nothing else going on — heart, lungs, kidneys, brain, the maintenance of tissue. It is measured after a twelve-hour fast, lying still, awake, in a thermally neutral room. Resting metabolic rate (RMR) is the same idea measured under slightly looser conditions, and runs a few per cent higher. Most calculators, this one included, use the terms interchangeably because the equations were fitted on a mix of both.
It is not your calorie target. For most people it is 60–75% of what they actually spend in a day; everything else is movement and digestion. To eat to it would be to ignore the rest.
Which equation to use
Mifflin–St Jeor (1990) is the default here and the usual recommendation. It was fitted on 498 adults and predicts measured resting rate within 10% more often than the alternatives.
Harris–Benedict dates to 1919 and was revised by Roza and Shizgal in 1984; this page uses the revision. It reads roughly 5% high for most people, which is why it has largely been retired. If a calculator gives you a noticeably bigger number than this page does, it is usually running the 1919 original.
Katch–McArdle ignores height and sex entirely and works from lean body mass, so it is the better equation — but only if your body fat figure is any good. Fed a number from a bathroom-scale bioimpedance reading, it inherits that error. Cunningham does the same thing from a sample of trained athletes, and reads higher as a result.
From BMR to maintenance, honestly
Multiply by an activity factor and you get maintenance. The factors are blunt — five tiers covering every life there is — and people reliably pick one too high. If you train three times a week and sit down the rest of it, you are "lightly active", not "moderately active". Choosing the tier above costs about 250 kcal a day of imaginary burn, which is most of a deficit.
The maintenance calculator does this step properly, and thedeficit calculator turns it into a daily target.
How accurate is this?
Every equation here is a regression fitted to a population. Roughly two-thirds of people fall within 10% of the prediction, which on a 1,700 kcal BMR is a 340 kcal spread. Thyroid conditions, some medication, and a long history of weight loss all shift it further. The practical answer is to take the number as a starting point and correct it against two to four weeks of real weight data — which is what an app can do and a calculator cannot.
Should you eat below your BMR?
As a sustained plan, no. It is possible to construct a deficit that puts intake under BMR, and the arithmetic will not stop you, but it tends to come with poor training quality, more lean tissue lost per kilogram, and an intake too low to hit basic micronutrient needs. If the only way to reach your pace is to eat under this number, the pace is wrong — or the burn side needs to do more of the work. Anyone being asked to eat that little for medical reasons should be supervised while they do it.
These equations are for adults. They are not valid during pregnancy or breastfeeding, and not for under-18s.